From 881adb85358309ea9c6f707394002719982ec607 Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Fri, 25 Jul 2008 01:48:29 -0700
Subject: proc: always do ->release

Current two-stage scheme of removing PDE emphasizes one bug in proc:

		open
				rmmod
				remove_proc_entry
		close

->release won't be called because ->proc_fops were cleared.  In simple
cases it's small memory leak.

For every ->open, ->release has to be done.  List of openers is introduced
which is traversed at remove_proc_entry() if neeeded.

Discussions with Al long ago (sigh).

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/proc/inode.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 70 insertions(+), 4 deletions(-)

(limited to 'fs/proc/inode.c')

diff --git a/fs/proc/inode.c b/fs/proc/inode.c
index b08d10017911..354c08485825 100644
--- a/fs/proc/inode.c
+++ b/fs/proc/inode.c
@@ -126,12 +126,17 @@ static const struct super_operations proc_sops = {
 	.remount_fs	= proc_remount,
 };
 
-static void pde_users_dec(struct proc_dir_entry *pde)
+static void __pde_users_dec(struct proc_dir_entry *pde)
 {
-	spin_lock(&pde->pde_unload_lock);
 	pde->pde_users--;
 	if (pde->pde_unload_completion && pde->pde_users == 0)
 		complete(pde->pde_unload_completion);
+}
+
+static void pde_users_dec(struct proc_dir_entry *pde)
+{
+	spin_lock(&pde->pde_unload_lock);
+	__pde_users_dec(pde);
 	spin_unlock(&pde->pde_unload_lock);
 }
 
@@ -318,36 +323,97 @@ static int proc_reg_open(struct inode *inode, struct file *file)
 	struct proc_dir_entry *pde = PDE(inode);
 	int rv = 0;
 	int (*open)(struct inode *, struct file *);
+	int (*release)(struct inode *, struct file *);
+	struct pde_opener *pdeo;
+
+	/*
+	 * What for, you ask? Well, we can have open, rmmod, remove_proc_entry
+	 * sequence. ->release won't be called because ->proc_fops will be
+	 * cleared. Depending on complexity of ->release, consequences vary.
+	 *
+	 * We can't wait for mercy when close will be done for real, it's
+	 * deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
+	 * by hand in remove_proc_entry(). For this, save opener's credentials
+	 * for later.
+	 */
+	pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
+	if (!pdeo)
+		return -ENOMEM;
 
 	spin_lock(&pde->pde_unload_lock);
 	if (!pde->proc_fops) {
 		spin_unlock(&pde->pde_unload_lock);
+		kfree(pdeo);
 		return rv;
 	}
 	pde->pde_users++;
 	open = pde->proc_fops->open;
+	release = pde->proc_fops->release;
 	spin_unlock(&pde->pde_unload_lock);
 
 	if (open)
 		rv = open(inode, file);
 
-	pde_users_dec(pde);
+	spin_lock(&pde->pde_unload_lock);
+	if (rv == 0 && release) {
+		/* To know what to release. */
+		pdeo->inode = inode;
+		pdeo->file = file;
+		/* Strictly for "too late" ->release in proc_reg_release(). */
+		pdeo->release = release;
+		list_add(&pdeo->lh, &pde->pde_openers);
+	} else
+		kfree(pdeo);
+	__pde_users_dec(pde);
+	spin_unlock(&pde->pde_unload_lock);
 	return rv;
 }
 
+static struct pde_opener *find_pde_opener(struct proc_dir_entry *pde,
+					struct inode *inode, struct file *file)
+{
+	struct pde_opener *pdeo;
+
+	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
+		if (pdeo->inode == inode && pdeo->file == file)
+			return pdeo;
+	}
+	return NULL;
+}
+
 static int proc_reg_release(struct inode *inode, struct file *file)
 {
 	struct proc_dir_entry *pde = PDE(inode);
 	int rv = 0;
 	int (*release)(struct inode *, struct file *);
+	struct pde_opener *pdeo;
 
 	spin_lock(&pde->pde_unload_lock);
+	pdeo = find_pde_opener(pde, inode, file);
 	if (!pde->proc_fops) {
-		spin_unlock(&pde->pde_unload_lock);
+		/*
+		 * Can't simply exit, __fput() will think that everything is OK,
+		 * and move on to freeing struct file. remove_proc_entry() will
+		 * find slacker in opener's list and will try to do non-trivial
+		 * things with struct file. Therefore, remove opener from list.
+		 *
+		 * But if opener is removed from list, who will ->release it?
+		 */
+		if (pdeo) {
+			list_del(&pdeo->lh);
+			spin_unlock(&pde->pde_unload_lock);
+			rv = pdeo->release(inode, file);
+			kfree(pdeo);
+		} else
+			spin_unlock(&pde->pde_unload_lock);
 		return rv;
 	}
 	pde->pde_users++;
 	release = pde->proc_fops->release;
+	if (pdeo) {
+		list_del(&pdeo->lh);
+		kfree(pdeo);
+	}
 	spin_unlock(&pde->pde_unload_lock);
 
 	if (release)
-- 
cgit v1.2.3


From a9bd4a3e070ba7494f154e1a11687a8a957d88dc Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Fri, 25 Jul 2008 01:48:30 -0700
Subject: proc: remove pathetic remount code

MS_RMT_MASK will unmask changes in do_remount_sb() anyway.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/proc/inode.c | 7 -------
 1 file changed, 7 deletions(-)

(limited to 'fs/proc/inode.c')

diff --git a/fs/proc/inode.c b/fs/proc/inode.c
index 354c08485825..02eca2ed9dd7 100644
--- a/fs/proc/inode.c
+++ b/fs/proc/inode.c
@@ -111,19 +111,12 @@ int __init proc_init_inodecache(void)
 	return 0;
 }
 
-static int proc_remount(struct super_block *sb, int *flags, char *data)
-{
-	*flags |= MS_NODIRATIME;
-	return 0;
-}
-
 static const struct super_operations proc_sops = {
 	.alloc_inode	= proc_alloc_inode,
 	.destroy_inode	= proc_destroy_inode,
 	.drop_inode	= generic_delete_inode,
 	.delete_inode	= proc_delete_inode,
 	.statfs		= simple_statfs,
-	.remount_fs	= proc_remount,
 };
 
 static void __pde_users_dec(struct proc_dir_entry *pde)
-- 
cgit v1.2.3


From 51cc50685a4275c6a02653670af9f108a64e01cf Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Fri, 25 Jul 2008 19:45:34 -0700
Subject: SL*B: drop kmem cache argument from constructor

Kmem cache passed to constructor is only needed for constructors that are
themselves multiplexeres.  Nobody uses this "feature", nor does anybody uses
passed kmem cache in non-trivial way, so pass only pointer to object.

Non-trivial places are:
	arch/powerpc/mm/init_64.c
	arch/powerpc/mm/hugetlbpage.c

This is flag day, yes.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Jon Tollefson <kniht@linux.vnet.ibm.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Matt Mackall <mpm@selenic.com>
[akpm@linux-foundation.org: fix arch/powerpc/mm/hugetlbpage.c]
[akpm@linux-foundation.org: fix mm/slab.c]
[akpm@linux-foundation.org: fix ubifs]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/arm/plat-s3c24xx/dma.c               |  2 +-
 arch/powerpc/kernel/rtas_flash.c          |  2 +-
 arch/powerpc/mm/hugetlbpage.c             |  9 ++-------
 arch/powerpc/mm/init_64.c                 | 24 +++++++++---------------
 arch/powerpc/platforms/cell/spufs/inode.c |  2 +-
 arch/sh/mm/pmb.c                          |  2 +-
 arch/xtensa/mm/init.c                     |  2 +-
 drivers/usb/mon/mon_text.c                |  4 ++--
 fs/adfs/super.c                           |  2 +-
 fs/affs/super.c                           |  2 +-
 fs/afs/super.c                            |  4 ++--
 fs/befs/linuxvfs.c                        |  2 +-
 fs/bfs/inode.c                            |  2 +-
 fs/block_dev.c                            |  2 +-
 fs/buffer.c                               |  2 +-
 fs/cifs/cifsfs.c                          |  2 +-
 fs/coda/inode.c                           |  2 +-
 fs/ecryptfs/main.c                        |  4 ++--
 fs/efs/super.c                            |  2 +-
 fs/ext2/super.c                           |  2 +-
 fs/ext3/super.c                           |  2 +-
 fs/ext4/super.c                           |  2 +-
 fs/fat/cache.c                            |  2 +-
 fs/fat/inode.c                            |  2 +-
 fs/fuse/inode.c                           |  2 +-
 fs/gfs2/main.c                            |  4 ++--
 fs/hfs/super.c                            |  2 +-
 fs/hfsplus/super.c                        |  2 +-
 fs/hpfs/super.c                           |  2 +-
 fs/hugetlbfs/inode.c                      |  2 +-
 fs/inode.c                                |  2 +-
 fs/isofs/inode.c                          |  2 +-
 fs/jffs2/super.c                          |  2 +-
 fs/jfs/jfs_metapage.c                     |  2 +-
 fs/jfs/super.c                            |  2 +-
 fs/locks.c                                |  2 +-
 fs/minix/inode.c                          |  2 +-
 fs/ncpfs/inode.c                          |  2 +-
 fs/nfs/inode.c                            |  2 +-
 fs/ntfs/super.c                           |  2 +-
 fs/ocfs2/dlm/dlmfs.c                      |  3 +--
 fs/ocfs2/super.c                          |  2 +-
 fs/openpromfs/inode.c                     |  2 +-
 fs/proc/inode.c                           |  2 +-
 fs/qnx4/inode.c                           |  2 +-
 fs/reiserfs/super.c                       |  2 +-
 fs/romfs/inode.c                          |  2 +-
 fs/smbfs/inode.c                          |  2 +-
 fs/sysv/inode.c                           |  2 +-
 fs/ubifs/super.c                          |  2 +-
 fs/udf/super.c                            |  2 +-
 fs/ufs/super.c                            |  2 +-
 fs/xfs/linux-2.6/kmem.h                   |  2 +-
 fs/xfs/linux-2.6/xfs_super.c              |  1 -
 include/linux/slab.h                      |  2 +-
 include/linux/slub_def.h                  |  2 +-
 ipc/mqueue.c                              |  2 +-
 kernel/fork.c                             |  2 +-
 lib/idr.c                                 |  2 +-
 lib/radix-tree.c                          |  2 +-
 mm/rmap.c                                 |  2 +-
 mm/shmem.c                                |  2 +-
 mm/slab.c                                 | 11 +++++------
 mm/slob.c                                 |  7 +++----
 mm/slub.c                                 | 13 ++++++-------
 net/socket.c                              |  2 +-
 net/sunrpc/rpc_pipe.c                     |  2 +-
 67 files changed, 90 insertions(+), 106 deletions(-)

(limited to 'fs/proc/inode.c')

diff --git a/arch/arm/plat-s3c24xx/dma.c b/arch/arm/plat-s3c24xx/dma.c
index 60f162dc4fad..8c5e656d5d8c 100644
--- a/arch/arm/plat-s3c24xx/dma.c
+++ b/arch/arm/plat-s3c24xx/dma.c
@@ -1304,7 +1304,7 @@ struct sysdev_class dma_sysclass = {
 
 /* kmem cache implementation */
 
-static void s3c2410_dma_cache_ctor(struct kmem_cache *c, void *p)
+static void s3c2410_dma_cache_ctor(void *p)
 {
 	memset(p, 0, sizeof(struct s3c2410_dma_buf));
 }
diff --git a/arch/powerpc/kernel/rtas_flash.c b/arch/powerpc/kernel/rtas_flash.c
index 09ded5c424a9..149cb112cd1a 100644
--- a/arch/powerpc/kernel/rtas_flash.c
+++ b/arch/powerpc/kernel/rtas_flash.c
@@ -286,7 +286,7 @@ static ssize_t rtas_flash_read(struct file *file, char __user *buf,
 }
 
 /* constructor for flash_block_cache */
-void rtas_block_ctor(struct kmem_cache *cache, void *ptr)
+void rtas_block_ctor(void *ptr)
 {
 	memset(ptr, 0, RTAS_BLK_SIZE);
 }
diff --git a/arch/powerpc/mm/hugetlbpage.c b/arch/powerpc/mm/hugetlbpage.c
index fb42c4dd3217..ed0aab0208a6 100644
--- a/arch/powerpc/mm/hugetlbpage.c
+++ b/arch/powerpc/mm/hugetlbpage.c
@@ -113,7 +113,7 @@ static inline pte_t *hugepte_offset(hugepd_t *hpdp, unsigned long addr,
 static int __hugepte_alloc(struct mm_struct *mm, hugepd_t *hpdp,
 			   unsigned long address, unsigned int psize)
 {
-	pte_t *new = kmem_cache_alloc(huge_pgtable_cache(psize),
+	pte_t *new = kmem_cache_zalloc(huge_pgtable_cache(psize),
 				      GFP_KERNEL|__GFP_REPEAT);
 
 	if (! new)
@@ -730,11 +730,6 @@ static int __init hugepage_setup_sz(char *str)
 }
 __setup("hugepagesz=", hugepage_setup_sz);
 
-static void zero_ctor(struct kmem_cache *cache, void *addr)
-{
-	memset(addr, 0, kmem_cache_size(cache));
-}
-
 static int __init hugetlbpage_init(void)
 {
 	unsigned int psize;
@@ -756,7 +751,7 @@ static int __init hugetlbpage_init(void)
 						HUGEPTE_TABLE_SIZE(psize),
 						HUGEPTE_TABLE_SIZE(psize),
 						0,
-						zero_ctor);
+						NULL);
 			if (!huge_pgtable_cache(psize))
 				panic("hugetlbpage_init(): could not create %s"\
 				      "\n", HUGEPTE_CACHE_NAME(psize));
diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c
index a41bc5aa2043..4f7df85129d8 100644
--- a/arch/powerpc/mm/init_64.c
+++ b/arch/powerpc/mm/init_64.c
@@ -136,9 +136,14 @@ static int __init setup_kcore(void)
 module_init(setup_kcore);
 #endif
 
-static void zero_ctor(struct kmem_cache *cache, void *addr)
+static void pgd_ctor(void *addr)
 {
-	memset(addr, 0, kmem_cache_size(cache));
+	memset(addr, 0, PGD_TABLE_SIZE);
+}
+
+static void pmd_ctor(void *addr)
+{
+	memset(addr, 0, PMD_TABLE_SIZE);
 }
 
 static const unsigned int pgtable_cache_size[2] = {
@@ -163,19 +168,8 @@ struct kmem_cache *pgtable_cache[ARRAY_SIZE(pgtable_cache_size)];
 
 void pgtable_cache_init(void)
 {
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(pgtable_cache_size); i++) {
-		int size = pgtable_cache_size[i];
-		const char *name = pgtable_cache_name[i];
-
-		pr_debug("Allocating page table cache %s (#%d) "
-			"for size: %08x...\n", name, i, size);
-		pgtable_cache[i] = kmem_cache_create(name,
-						     size, size,
-						     SLAB_PANIC,
-						     zero_ctor);
-	}
+	pgtable_cache[0] = kmem_cache_create(pgtable_cache_name[0], PGD_TABLE_SIZE, PGD_TABLE_SIZE, SLAB_PANIC, pgd_ctor);
+	pgtable_cache[1] = kmem_cache_create(pgtable_cache_name[1], PMD_TABLE_SIZE, PMD_TABLE_SIZE, SLAB_PANIC, pmd_ctor);
 }
 
 #ifdef CONFIG_SPARSEMEM_VMEMMAP
diff --git a/arch/powerpc/platforms/cell/spufs/inode.c b/arch/powerpc/platforms/cell/spufs/inode.c
index 7123472801d9..690ca7b0dcf6 100644
--- a/arch/powerpc/platforms/cell/spufs/inode.c
+++ b/arch/powerpc/platforms/cell/spufs/inode.c
@@ -78,7 +78,7 @@ spufs_destroy_inode(struct inode *inode)
 }
 
 static void
-spufs_init_once(struct kmem_cache *cachep, void *p)
+spufs_init_once(void *p)
 {
 	struct spufs_inode_info *ei = p;
 
diff --git a/arch/sh/mm/pmb.c b/arch/sh/mm/pmb.c
index 0b0ec6e04753..46911bcbf17b 100644
--- a/arch/sh/mm/pmb.c
+++ b/arch/sh/mm/pmb.c
@@ -293,7 +293,7 @@ void pmb_unmap(unsigned long addr)
 	} while (pmbe);
 }
 
-static void pmb_cache_ctor(struct kmem_cache *cachep, void *pmb)
+static void pmb_cache_ctor(void *pmb)
 {
 	struct pmb_entry *pmbe = pmb;
 
diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c
index 81d0560eaea2..ee261005b363 100644
--- a/arch/xtensa/mm/init.c
+++ b/arch/xtensa/mm/init.c
@@ -309,7 +309,7 @@ void show_mem(void)
 
 struct kmem_cache *pgtable_cache __read_mostly;
 
-static void pgd_ctor(struct kmem_cache *cache, void* addr)
+static void pgd_ctor(void* addr)
 {
 	pte_t* ptep = (pte_t*)addr;
 	int i;
diff --git a/drivers/usb/mon/mon_text.c b/drivers/usb/mon/mon_text.c
index 5e3e4e9b6c77..1f715436d6d3 100644
--- a/drivers/usb/mon/mon_text.c
+++ b/drivers/usb/mon/mon_text.c
@@ -87,7 +87,7 @@ struct mon_reader_text {
 
 static struct dentry *mon_dir;		/* Usually /sys/kernel/debug/usbmon */
 
-static void mon_text_ctor(struct kmem_cache *, void *);
+static void mon_text_ctor(void *);
 
 struct mon_text_ptr {
 	int cnt, limit;
@@ -720,7 +720,7 @@ void mon_text_del(struct mon_bus *mbus)
 /*
  * Slab interface: constructor.
  */
-static void mon_text_ctor(struct kmem_cache *slab, void *mem)
+static void mon_text_ctor(void *mem)
 {
 	/*
 	 * Nothing to initialize. No, really!
diff --git a/fs/adfs/super.c b/fs/adfs/super.c
index 9e421eeb672b..26f3b43726bb 100644
--- a/fs/adfs/super.c
+++ b/fs/adfs/super.c
@@ -249,7 +249,7 @@ static void adfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(adfs_inode_cachep, ADFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct adfs_inode_info *ei = (struct adfs_inode_info *) foo;
 
diff --git a/fs/affs/super.c b/fs/affs/super.c
index 4e0309566406..3a89094f93d0 100644
--- a/fs/affs/super.c
+++ b/fs/affs/super.c
@@ -90,7 +90,7 @@ static void affs_destroy_inode(struct inode *inode)
 	kmem_cache_free(affs_inode_cachep, AFFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct affs_inode_info *ei = (struct affs_inode_info *) foo;
 
diff --git a/fs/afs/super.c b/fs/afs/super.c
index 7e3faeef6818..250d8c4d66e4 100644
--- a/fs/afs/super.c
+++ b/fs/afs/super.c
@@ -27,7 +27,7 @@
 
 #define AFS_FS_MAGIC 0x6B414653 /* 'kAFS' */
 
-static void afs_i_init_once(struct kmem_cache *cachep, void *foo);
+static void afs_i_init_once(void *foo);
 static int afs_get_sb(struct file_system_type *fs_type,
 		      int flags, const char *dev_name,
 		      void *data, struct vfsmount *mnt);
@@ -449,7 +449,7 @@ static void afs_put_super(struct super_block *sb)
 /*
  * initialise an inode cache slab element prior to any use
  */
-static void afs_i_init_once(struct kmem_cache *cachep, void *_vnode)
+static void afs_i_init_once(void *_vnode)
 {
 	struct afs_vnode *vnode = _vnode;
 
diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c
index e8717de3bab3..02c6e62b72f8 100644
--- a/fs/befs/linuxvfs.c
+++ b/fs/befs/linuxvfs.c
@@ -289,7 +289,7 @@ befs_destroy_inode(struct inode *inode)
         kmem_cache_free(befs_inode_cachep, BEFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
         struct befs_inode_info *bi = (struct befs_inode_info *) foo;
 
diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c
index 053e690ec9ed..0ed57b5ee012 100644
--- a/fs/bfs/inode.c
+++ b/fs/bfs/inode.c
@@ -264,7 +264,7 @@ static void bfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(bfs_inode_cachep, BFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct bfs_inode_info *bi = foo;
 
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 10d8a0aa871a..dcf37cada369 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -271,7 +271,7 @@ static void bdev_destroy_inode(struct inode *inode)
 	kmem_cache_free(bdev_cachep, bdi);
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct bdev_inode *ei = (struct bdev_inode *) foo;
 	struct block_device *bdev = &ei->bdev;
diff --git a/fs/buffer.c b/fs/buffer.c
index 109b261192d9..5fd497cdd6f3 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -3272,7 +3272,7 @@ int bh_submit_read(struct buffer_head *bh)
 EXPORT_SYMBOL(bh_submit_read);
 
 static void
-init_buffer_head(struct kmem_cache *cachep, void *data)
+init_buffer_head(void *data)
 {
 	struct buffer_head *bh = data;
 
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index 22857c639df5..fe5f6809cba6 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -766,7 +766,7 @@ const struct file_operations cifs_dir_ops = {
 };
 
 static void
-cifs_init_once(struct kmem_cache *cachep, void *inode)
+cifs_init_once(void *inode)
 {
 	struct cifsInodeInfo *cifsi = inode;
 
diff --git a/fs/coda/inode.c b/fs/coda/inode.c
index 2f58dfc70083..830f51abb971 100644
--- a/fs/coda/inode.c
+++ b/fs/coda/inode.c
@@ -58,7 +58,7 @@ static void coda_destroy_inode(struct inode *inode)
 	kmem_cache_free(coda_inode_cachep, ITOC(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct coda_inode_info *ei = (struct coda_inode_info *) foo;
 
diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c
index 6f403cfba14f..448dfd597b5f 100644
--- a/fs/ecryptfs/main.c
+++ b/fs/ecryptfs/main.c
@@ -578,7 +578,7 @@ static struct file_system_type ecryptfs_fs_type = {
  * Initializes the ecryptfs_inode_info_cache when it is created
  */
 static void
-inode_info_init_once(struct kmem_cache *cachep, void *vptr)
+inode_info_init_once(void *vptr)
 {
 	struct ecryptfs_inode_info *ei = (struct ecryptfs_inode_info *)vptr;
 
@@ -589,7 +589,7 @@ static struct ecryptfs_cache_info {
 	struct kmem_cache **cache;
 	const char *name;
 	size_t size;
-	void (*ctor)(struct kmem_cache *cache, void *obj);
+	void (*ctor)(void *obj);
 } ecryptfs_cache_infos[] = {
 	{
 		.cache = &ecryptfs_auth_tok_list_item_cache,
diff --git a/fs/efs/super.c b/fs/efs/super.c
index d733531b55e2..567b134fa1f1 100644
--- a/fs/efs/super.c
+++ b/fs/efs/super.c
@@ -70,7 +70,7 @@ static void efs_destroy_inode(struct inode *inode)
 	kmem_cache_free(efs_inode_cachep, INODE_INFO(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct efs_inode_info *ei = (struct efs_inode_info *) foo;
 
diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 31308a3b0b8b..fd88c7b43e66 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -159,7 +159,7 @@ static void ext2_destroy_inode(struct inode *inode)
 	kmem_cache_free(ext2_inode_cachep, EXT2_I(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct ext2_inode_info *ei = (struct ext2_inode_info *) foo;
 
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index 615788c6843a..8ddced384674 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -472,7 +472,7 @@ static void ext3_destroy_inode(struct inode *inode)
 	kmem_cache_free(ext3_inode_cachep, EXT3_I(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct ext3_inode_info *ei = (struct ext3_inode_info *) foo;
 
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 1cb371dcd609..b5479b1dff14 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -595,7 +595,7 @@ static void ext4_destroy_inode(struct inode *inode)
 	kmem_cache_free(ext4_inode_cachep, EXT4_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct ext4_inode_info *ei = (struct ext4_inode_info *) foo;
 
diff --git a/fs/fat/cache.c b/fs/fat/cache.c
index 3a9ecac8d61f..3222f51c41cf 100644
--- a/fs/fat/cache.c
+++ b/fs/fat/cache.c
@@ -36,7 +36,7 @@ static inline int fat_max_cache(struct inode *inode)
 
 static struct kmem_cache *fat_cache_cachep;
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct fat_cache *cache = (struct fat_cache *)foo;
 
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index 23676f9d79ce..6d266d793e2c 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -498,7 +498,7 @@ static void fat_destroy_inode(struct inode *inode)
 	kmem_cache_free(fat_inode_cachep, MSDOS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct msdos_inode_info *ei = (struct msdos_inode_info *)foo;
 
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index 7d2f7d6e22e2..d2249f174e20 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -956,7 +956,7 @@ static inline void unregister_fuseblk(void)
 }
 #endif
 
-static void fuse_inode_init_once(struct kmem_cache *cachep, void *foo)
+static void fuse_inode_init_once(void *foo)
 {
 	struct inode * inode = foo;
 
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index bcc668d0fadd..bb2cc303ac29 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -24,7 +24,7 @@
 #include "util.h"
 #include "glock.h"
 
-static void gfs2_init_inode_once(struct kmem_cache *cachep, void *foo)
+static void gfs2_init_inode_once(void *foo)
 {
 	struct gfs2_inode *ip = foo;
 
@@ -33,7 +33,7 @@ static void gfs2_init_inode_once(struct kmem_cache *cachep, void *foo)
 	ip->i_alloc = NULL;
 }
 
-static void gfs2_init_glock_once(struct kmem_cache *cachep, void *foo)
+static void gfs2_init_glock_once(void *foo)
 {
 	struct gfs2_glock *gl = foo;
 
diff --git a/fs/hfs/super.c b/fs/hfs/super.c
index ac2ec5ef66e4..4abb1047c689 100644
--- a/fs/hfs/super.c
+++ b/fs/hfs/super.c
@@ -432,7 +432,7 @@ static struct file_system_type hfs_fs_type = {
 	.fs_flags	= FS_REQUIRES_DEV,
 };
 
-static void hfs_init_once(struct kmem_cache *cachep, void *p)
+static void hfs_init_once(void *p)
 {
 	struct hfs_inode_info *i = p;
 
diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c
index 3859118531c7..e834e578c93f 100644
--- a/fs/hfsplus/super.c
+++ b/fs/hfsplus/super.c
@@ -485,7 +485,7 @@ static struct file_system_type hfsplus_fs_type = {
 	.fs_flags	= FS_REQUIRES_DEV,
 };
 
-static void hfsplus_init_once(struct kmem_cache *cachep, void *p)
+static void hfsplus_init_once(void *p)
 {
 	struct hfsplus_inode_info *i = p;
 
diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c
index f63a699ec659..b8ae9c90ada0 100644
--- a/fs/hpfs/super.c
+++ b/fs/hpfs/super.c
@@ -173,7 +173,7 @@ static void hpfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(hpfs_inode_cachep, hpfs_i(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct hpfs_inode_info *ei = (struct hpfs_inode_info *) foo;
 
diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index dbd01d262ca4..3f58923fb39b 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -705,7 +705,7 @@ static const struct address_space_operations hugetlbfs_aops = {
 };
 
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct hugetlbfs_inode_info *ei = (struct hugetlbfs_inode_info *)foo;
 
diff --git a/fs/inode.c b/fs/inode.c
index 35b6414522ea..b6726f644530 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -224,7 +224,7 @@ void inode_init_once(struct inode *inode)
 
 EXPORT_SYMBOL(inode_init_once);
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct inode * inode = (struct inode *) foo;
 
diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c
index 044a254d526b..26948a6033b6 100644
--- a/fs/isofs/inode.c
+++ b/fs/isofs/inode.c
@@ -73,7 +73,7 @@ static void isofs_destroy_inode(struct inode *inode)
 	kmem_cache_free(isofs_inode_cachep, ISOFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct iso_inode_info *ei = foo;
 
diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c
index 7da69eae49e4..efd401257ed9 100644
--- a/fs/jffs2/super.c
+++ b/fs/jffs2/super.c
@@ -44,7 +44,7 @@ static void jffs2_destroy_inode(struct inode *inode)
 	kmem_cache_free(jffs2_inode_cachep, JFFS2_INODE_INFO(inode));
 }
 
-static void jffs2_i_init_once(struct kmem_cache *cachep, void *foo)
+static void jffs2_i_init_once(void *foo)
 {
 	struct jffs2_inode_info *f = foo;
 
diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c
index 854ff0ec574f..c350057087dd 100644
--- a/fs/jfs/jfs_metapage.c
+++ b/fs/jfs/jfs_metapage.c
@@ -182,7 +182,7 @@ static inline void remove_metapage(struct page *page, struct metapage *mp)
 
 #endif
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct metapage *mp = (struct metapage *)foo;
 
diff --git a/fs/jfs/super.c b/fs/jfs/super.c
index 359c091d8965..3630718be395 100644
--- a/fs/jfs/super.c
+++ b/fs/jfs/super.c
@@ -760,7 +760,7 @@ static struct file_system_type jfs_fs_type = {
 	.fs_flags	= FS_REQUIRES_DEV,
 };
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct jfs_inode_info *jfs_ip = (struct jfs_inode_info *) foo;
 
diff --git a/fs/locks.c b/fs/locks.c
index 01490300f7cb..5eb259e3cd38 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -201,7 +201,7 @@ EXPORT_SYMBOL(locks_init_lock);
  * Initialises the fields of the file lock which are invariant for
  * free file_locks.
  */
-static void init_once(struct kmem_cache *cache, void *foo)
+static void init_once(void *foo)
 {
 	struct file_lock *lock = (struct file_lock *) foo;
 
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 523d73713418..d1d1eb84679d 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -68,7 +68,7 @@ static void minix_destroy_inode(struct inode *inode)
 	kmem_cache_free(minix_inode_cachep, minix_i(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct minix_inode_info *ei = (struct minix_inode_info *) foo;
 
diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c
index 2e5ab1204dec..d642f0e5b365 100644
--- a/fs/ncpfs/inode.c
+++ b/fs/ncpfs/inode.c
@@ -64,7 +64,7 @@ static void ncp_destroy_inode(struct inode *inode)
 	kmem_cache_free(ncp_inode_cachep, NCP_FINFO(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct ncp_inode_info *ei = (struct ncp_inode_info *) foo;
 
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index df23f987da6b..52daefa2f521 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1242,7 +1242,7 @@ static inline void nfs4_init_once(struct nfs_inode *nfsi)
 #endif
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct nfs_inode *nfsi = (struct nfs_inode *) foo;
 
diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c
index 3e76f3b216bc..4a46743b5077 100644
--- a/fs/ntfs/super.c
+++ b/fs/ntfs/super.c
@@ -3080,7 +3080,7 @@ struct kmem_cache *ntfs_inode_cache;
 struct kmem_cache *ntfs_big_inode_cache;
 
 /* Init once constructor for the inode slab cache. */
-static void ntfs_big_inode_init_once(struct kmem_cache *cachep, void *foo)
+static void ntfs_big_inode_init_once(void *foo)
 {
 	ntfs_inode *ni = (ntfs_inode *)foo;
 
diff --git a/fs/ocfs2/dlm/dlmfs.c b/fs/ocfs2/dlm/dlmfs.c
index e48aba698b77..533a789c3ef8 100644
--- a/fs/ocfs2/dlm/dlmfs.c
+++ b/fs/ocfs2/dlm/dlmfs.c
@@ -267,8 +267,7 @@ static ssize_t dlmfs_file_write(struct file *filp,
 	return writelen;
 }
 
-static void dlmfs_init_once(struct kmem_cache *cachep,
-			    void *foo)
+static void dlmfs_init_once(void *foo)
 {
 	struct dlmfs_inode_private *ip =
 		(struct dlmfs_inode_private *) foo;
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index ccecfe5094fa..2560b33889aa 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -1118,7 +1118,7 @@ bail:
 	return status;
 }
 
-static void ocfs2_inode_init_once(struct kmem_cache *cachep, void *data)
+static void ocfs2_inode_init_once(void *data)
 {
 	struct ocfs2_inode_info *oi = data;
 
diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c
index d17b4fd204e1..9f5b054f06b9 100644
--- a/fs/openpromfs/inode.c
+++ b/fs/openpromfs/inode.c
@@ -430,7 +430,7 @@ static struct file_system_type openprom_fs_type = {
 	.kill_sb	= kill_anon_super,
 };
 
-static void op_inode_init_once(struct kmem_cache * cachep, void *data)
+static void op_inode_init_once(void *data)
 {
 	struct op_inode_info *oi = (struct op_inode_info *) data;
 
diff --git a/fs/proc/inode.c b/fs/proc/inode.c
index 02eca2ed9dd7..b37f25dc45a5 100644
--- a/fs/proc/inode.c
+++ b/fs/proc/inode.c
@@ -94,7 +94,7 @@ static void proc_destroy_inode(struct inode *inode)
 	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct proc_inode *ei = (struct proc_inode *) foo;
 
diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c
index b31ab78052b3..2aad1044b84c 100644
--- a/fs/qnx4/inode.c
+++ b/fs/qnx4/inode.c
@@ -553,7 +553,7 @@ static void qnx4_destroy_inode(struct inode *inode)
 	kmem_cache_free(qnx4_inode_cachep, qnx4_i(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct qnx4_inode_info *ei = (struct qnx4_inode_info *) foo;
 
diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index 2ec748ba0bd3..879e54d35c2d 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -521,7 +521,7 @@ static void reiserfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(reiserfs_inode_cachep, REISERFS_I(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct reiserfs_inode_info *ei = (struct reiserfs_inode_info *)foo;
 
diff --git a/fs/romfs/inode.c b/fs/romfs/inode.c
index 3f13d491c7c7..8e51a2aaa977 100644
--- a/fs/romfs/inode.c
+++ b/fs/romfs/inode.c
@@ -577,7 +577,7 @@ static void romfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(romfs_inode_cachep, ROMFS_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct romfs_inode_info *ei = foo;
 
diff --git a/fs/smbfs/inode.c b/fs/smbfs/inode.c
index 376ef3ee6ed7..3528f40ffb0f 100644
--- a/fs/smbfs/inode.c
+++ b/fs/smbfs/inode.c
@@ -67,7 +67,7 @@ static void smb_destroy_inode(struct inode *inode)
 	kmem_cache_free(smb_inode_cachep, SMB_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct smb_inode_info *ei = (struct smb_inode_info *) foo;
 
diff --git a/fs/sysv/inode.c b/fs/sysv/inode.c
index c5d60de0658f..df0d435baa48 100644
--- a/fs/sysv/inode.c
+++ b/fs/sysv/inode.c
@@ -326,7 +326,7 @@ static void sysv_destroy_inode(struct inode *inode)
 	kmem_cache_free(sysv_inode_cachep, SYSV_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *p)
+static void init_once(void *p)
 {
 	struct sysv_inode_info *si = (struct sysv_inode_info *)p;
 
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index 00eb9c68ad03..ca1e2d4e03cc 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -1841,7 +1841,7 @@ static struct file_system_type ubifs_fs_type = {
 /*
  * Inode slab cache constructor.
  */
-static void inode_slab_ctor(struct kmem_cache *cachep, void *obj)
+static void inode_slab_ctor(void *obj)
 {
 	struct ubifs_inode *ui = obj;
 	inode_init_once(&ui->vfs_inode);
diff --git a/fs/udf/super.c b/fs/udf/super.c
index 44cc702f96cc..5698bbf83bbf 100644
--- a/fs/udf/super.c
+++ b/fs/udf/super.c
@@ -148,7 +148,7 @@ static void udf_destroy_inode(struct inode *inode)
 	kmem_cache_free(udf_inode_cachep, UDF_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct udf_inode_info *ei = (struct udf_inode_info *)foo;
 
diff --git a/fs/ufs/super.c b/fs/ufs/super.c
index 227c9d700040..3e30e40aa24d 100644
--- a/fs/ufs/super.c
+++ b/fs/ufs/super.c
@@ -1302,7 +1302,7 @@ static void ufs_destroy_inode(struct inode *inode)
 	kmem_cache_free(ufs_inode_cachep, UFS_I(inode));
 }
 
-static void init_once(struct kmem_cache * cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct ufs_inode_info *ei = (struct ufs_inode_info *) foo;
 
diff --git a/fs/xfs/linux-2.6/kmem.h b/fs/xfs/linux-2.6/kmem.h
index 5e9564902976..a20683cf74dd 100644
--- a/fs/xfs/linux-2.6/kmem.h
+++ b/fs/xfs/linux-2.6/kmem.h
@@ -79,7 +79,7 @@ kmem_zone_init(int size, char *zone_name)
 
 static inline kmem_zone_t *
 kmem_zone_init_flags(int size, char *zone_name, unsigned long flags,
-		     void (*construct)(kmem_zone_t *, void *))
+		     void (*construct)(void *))
 {
 	return kmem_cache_create(zone_name, size, 0, flags, construct);
 }
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 742b2c7852c1..943381284e2e 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -843,7 +843,6 @@ xfs_fs_destroy_inode(
 
 STATIC void
 xfs_fs_inode_init_once(
-	kmem_zone_t		*zonep,
 	void			*vnode)
 {
 	inode_init_once(vn_to_inode((bhv_vnode_t *)vnode));
diff --git a/include/linux/slab.h b/include/linux/slab.h
index 41103910f8a2..9ff8e8499403 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -58,7 +58,7 @@ int slab_is_available(void);
 
 struct kmem_cache *kmem_cache_create(const char *, size_t, size_t,
 			unsigned long,
-			void (*)(struct kmem_cache *, void *));
+			void (*)(void *));
 void kmem_cache_destroy(struct kmem_cache *);
 int kmem_cache_shrink(struct kmem_cache *);
 void kmem_cache_free(struct kmem_cache *, void *);
diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h
index d117ea2825a9..5bad61a93f65 100644
--- a/include/linux/slub_def.h
+++ b/include/linux/slub_def.h
@@ -85,7 +85,7 @@ struct kmem_cache {
 	struct kmem_cache_order_objects min;
 	gfp_t allocflags;	/* gfp flags to use on each alloc */
 	int refcount;		/* Refcount for slab cache destroy */
-	void (*ctor)(struct kmem_cache *, void *);
+	void (*ctor)(void *);
 	int inuse;		/* Offset to metadata */
 	int align;		/* Alignment */
 	const char *name;	/* Name (only for display!) */
diff --git a/ipc/mqueue.c b/ipc/mqueue.c
index 1fdc2eb2f6d8..474984f9e032 100644
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -207,7 +207,7 @@ static int mqueue_get_sb(struct file_system_type *fs_type,
 	return get_sb_single(fs_type, flags, data, mqueue_fill_super, mnt);
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct mqueue_inode_info *p = (struct mqueue_inode_info *) foo;
 
diff --git a/kernel/fork.c b/kernel/fork.c
index b99d73e971a4..80e83e459b17 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1442,7 +1442,7 @@ long do_fork(unsigned long clone_flags,
 #define ARCH_MIN_MMSTRUCT_ALIGN 0
 #endif
 
-static void sighand_ctor(struct kmem_cache *cachep, void *data)
+static void sighand_ctor(void *data)
 {
 	struct sighand_struct *sighand = data;
 
diff --git a/lib/idr.c b/lib/idr.c
index 3476f8203e97..e728c7fccc4d 100644
--- a/lib/idr.c
+++ b/lib/idr.c
@@ -607,7 +607,7 @@ void *idr_replace(struct idr *idp, void *ptr, int id)
 }
 EXPORT_SYMBOL(idr_replace);
 
-static void idr_cache_ctor(struct kmem_cache *idr_layer_cache, void *idr_layer)
+static void idr_cache_ctor(void *idr_layer)
 {
 	memset(idr_layer, 0, sizeof(struct idr_layer));
 }
diff --git a/lib/radix-tree.c b/lib/radix-tree.c
index 9c4f1ffa2864..be86b32bc874 100644
--- a/lib/radix-tree.c
+++ b/lib/radix-tree.c
@@ -1183,7 +1183,7 @@ int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag)
 EXPORT_SYMBOL(radix_tree_tagged);
 
 static void
-radix_tree_node_ctor(struct kmem_cache *cachep, void *node)
+radix_tree_node_ctor(void *node)
 {
 	memset(node, 0, sizeof(struct radix_tree_node));
 }
diff --git a/mm/rmap.c b/mm/rmap.c
index abbd29f7c43f..39ae5a9bf382 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -138,7 +138,7 @@ void anon_vma_unlink(struct vm_area_struct *vma)
 		anon_vma_free(anon_vma);
 }
 
-static void anon_vma_ctor(struct kmem_cache *cachep, void *data)
+static void anon_vma_ctor(void *data)
 {
 	struct anon_vma *anon_vma = data;
 
diff --git a/mm/shmem.c b/mm/shmem.c
index 1089092aecaf..952d361774bb 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2352,7 +2352,7 @@ static void shmem_destroy_inode(struct inode *inode)
 	kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct shmem_inode_info *p = (struct shmem_inode_info *) foo;
 
diff --git a/mm/slab.c b/mm/slab.c
index 052e7d64537e..918f04f7fef1 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -406,7 +406,7 @@ struct kmem_cache {
 	unsigned int dflags;		/* dynamic flags */
 
 	/* constructor func */
-	void (*ctor)(struct kmem_cache *, void *);
+	void (*ctor)(void *obj);
 
 /* 5) cache creation/removal */
 	const char *name;
@@ -2137,8 +2137,7 @@ static int __init_refok setup_cpu_cache(struct kmem_cache *cachep)
  */
 struct kmem_cache *
 kmem_cache_create (const char *name, size_t size, size_t align,
-	unsigned long flags,
-	void (*ctor)(struct kmem_cache *, void *))
+	unsigned long flags, void (*ctor)(void *))
 {
 	size_t left_over, slab_size, ralign;
 	struct kmem_cache *cachep = NULL, *pc;
@@ -2653,7 +2652,7 @@ static void cache_init_objs(struct kmem_cache *cachep,
 		 * They must also be threaded.
 		 */
 		if (cachep->ctor && !(cachep->flags & SLAB_POISON))
-			cachep->ctor(cachep, objp + obj_offset(cachep));
+			cachep->ctor(objp + obj_offset(cachep));
 
 		if (cachep->flags & SLAB_RED_ZONE) {
 			if (*dbg_redzone2(cachep, objp) != RED_INACTIVE)
@@ -2669,7 +2668,7 @@ static void cache_init_objs(struct kmem_cache *cachep,
 					 cachep->buffer_size / PAGE_SIZE, 0);
 #else
 		if (cachep->ctor)
-			cachep->ctor(cachep, objp);
+			cachep->ctor(objp);
 #endif
 		slab_bufctl(slabp)[i] = i + 1;
 	}
@@ -3093,7 +3092,7 @@ static void *cache_alloc_debugcheck_after(struct kmem_cache *cachep,
 #endif
 	objp += obj_offset(cachep);
 	if (cachep->ctor && cachep->flags & SLAB_POISON)
-		cachep->ctor(cachep, objp);
+		cachep->ctor(objp);
 #if ARCH_SLAB_MINALIGN
 	if ((u32)objp & (ARCH_SLAB_MINALIGN-1)) {
 		printk(KERN_ERR "0x%p: not aligned to ARCH_SLAB_MINALIGN=%d\n",
diff --git a/mm/slob.c b/mm/slob.c
index de268eb7ac70..d8fbd4d1bfa7 100644
--- a/mm/slob.c
+++ b/mm/slob.c
@@ -525,12 +525,11 @@ struct kmem_cache {
 	unsigned int size, align;
 	unsigned long flags;
 	const char *name;
-	void (*ctor)(struct kmem_cache *, void *);
+	void (*ctor)(void *);
 };
 
 struct kmem_cache *kmem_cache_create(const char *name, size_t size,
-	size_t align, unsigned long flags,
-	void (*ctor)(struct kmem_cache *, void *))
+	size_t align, unsigned long flags, void (*ctor)(void *))
 {
 	struct kmem_cache *c;
 
@@ -575,7 +574,7 @@ void *kmem_cache_alloc_node(struct kmem_cache *c, gfp_t flags, int node)
 		b = slob_new_page(flags, get_order(c->size), node);
 
 	if (c->ctor)
-		c->ctor(c, b);
+		c->ctor(b);
 
 	return b;
 }
diff --git a/mm/slub.c b/mm/slub.c
index 77c21cf53ff9..b7e2cd5d82db 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -1012,7 +1012,7 @@ __setup("slub_debug", setup_slub_debug);
 
 static unsigned long kmem_cache_flags(unsigned long objsize,
 	unsigned long flags, const char *name,
-	void (*ctor)(struct kmem_cache *, void *))
+	void (*ctor)(void *))
 {
 	/*
 	 * Enable debugging if selected on the kernel commandline.
@@ -1040,7 +1040,7 @@ static inline int check_object(struct kmem_cache *s, struct page *page,
 static inline void add_full(struct kmem_cache_node *n, struct page *page) {}
 static inline unsigned long kmem_cache_flags(unsigned long objsize,
 	unsigned long flags, const char *name,
-	void (*ctor)(struct kmem_cache *, void *))
+	void (*ctor)(void *))
 {
 	return flags;
 }
@@ -1103,7 +1103,7 @@ static void setup_object(struct kmem_cache *s, struct page *page,
 {
 	setup_object_debug(s, page, object);
 	if (unlikely(s->ctor))
-		s->ctor(s, object);
+		s->ctor(object);
 }
 
 static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
@@ -2286,7 +2286,7 @@ static int calculate_sizes(struct kmem_cache *s, int forced_order)
 static int kmem_cache_open(struct kmem_cache *s, gfp_t gfpflags,
 		const char *name, size_t size,
 		size_t align, unsigned long flags,
-		void (*ctor)(struct kmem_cache *, void *))
+		void (*ctor)(void *))
 {
 	memset(s, 0, kmem_size);
 	s->name = name;
@@ -3042,7 +3042,7 @@ static int slab_unmergeable(struct kmem_cache *s)
 
 static struct kmem_cache *find_mergeable(size_t size,
 		size_t align, unsigned long flags, const char *name,
-		void (*ctor)(struct kmem_cache *, void *))
+		void (*ctor)(void *))
 {
 	struct kmem_cache *s;
 
@@ -3082,8 +3082,7 @@ static struct kmem_cache *find_mergeable(size_t size,
 }
 
 struct kmem_cache *kmem_cache_create(const char *name, size_t size,
-		size_t align, unsigned long flags,
-		void (*ctor)(struct kmem_cache *, void *))
+		size_t align, unsigned long flags, void (*ctor)(void *))
 {
 	struct kmem_cache *s;
 
diff --git a/net/socket.c b/net/socket.c
index 1310a82cbba7..8ef8ba81b9e2 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -265,7 +265,7 @@ static void sock_destroy_inode(struct inode *inode)
 			container_of(inode, struct socket_alloc, vfs_inode));
 }
 
-static void init_once(struct kmem_cache *cachep, void *foo)
+static void init_once(void *foo)
 {
 	struct socket_alloc *ei = (struct socket_alloc *)foo;
 
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 5a9b0e7828cd..23a2b8f6dc49 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -897,7 +897,7 @@ static struct file_system_type rpc_pipe_fs_type = {
 };
 
 static void
-init_once(struct kmem_cache * cachep, void *foo)
+init_once(void *foo)
 {
 	struct rpc_inode *rpci = (struct rpc_inode *) foo;
 
-- 
cgit v1.2.3


From 9043476f726802f4b00c96d0c4f418dde48d1304 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Tue, 15 Jul 2008 08:54:06 -0400
Subject: [PATCH] sanitize proc_sysctl

* keep references to ctl_table_head and ctl_table in /proc/sys inodes
* grab the former during operations, use the latter for access to
  entry if that succeeds
* have ->d_compare() check if table should be seen for one who does lookup;
  that allows us to avoid flipping inodes - if we have the same name resolve
  to different things, we'll just keep several dentries and ->d_compare()
  will reject the wrong ones.
* have ->lookup() and ->readdir() scan the table of our inode first, then
  walk all ctl_table_header and scan ->attached_by for those that are
  attached to our directory.
* implement ->getattr().
* get rid of insane amounts of tree-walking
* get rid of the need to know dentry in ->permission() and of the contortions
  induced by that.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/proc/inode.c         |   5 +
 fs/proc/proc_sysctl.c   | 427 ++++++++++++++++++++++--------------------------
 include/linux/proc_fs.h |   5 +
 include/linux/sysctl.h  |   1 +
 kernel/sysctl.c         |  15 ++
 5 files changed, 218 insertions(+), 235 deletions(-)

(limited to 'fs/proc/inode.c')

diff --git a/fs/proc/inode.c b/fs/proc/inode.c
index b37f25dc45a5..8bb03f056c28 100644
--- a/fs/proc/inode.c
+++ b/fs/proc/inode.c
@@ -17,6 +17,7 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/smp_lock.h>
+#include <linux/sysctl.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -65,6 +66,8 @@ static void proc_delete_inode(struct inode *inode)
 			module_put(de->owner);
 		de_put(de);
 	}
+	if (PROC_I(inode)->sysctl)
+		sysctl_head_put(PROC_I(inode)->sysctl);
 	clear_inode(inode);
 }
 
@@ -84,6 +87,8 @@ static struct inode *proc_alloc_inode(struct super_block *sb)
 	ei->fd = 0;
 	ei->op.proc_get_link = NULL;
 	ei->pde = NULL;
+	ei->sysctl = NULL;
+	ei->sysctl_entry = NULL;
 	inode = &ei->vfs_inode;
 	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
 	return inode;
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index 5acc001d49f6..fa1ec2433e44 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -10,149 +10,110 @@
 static struct dentry_operations proc_sys_dentry_operations;
 static const struct file_operations proc_sys_file_operations;
 static const struct inode_operations proc_sys_inode_operations;
+static const struct file_operations proc_sys_dir_file_operations;
+static const struct inode_operations proc_sys_dir_operations;
 
-static void proc_sys_refresh_inode(struct inode *inode, struct ctl_table *table)
-{
-	/* Refresh the cached information bits in the inode */
-	if (table) {
-		inode->i_uid = 0;
-		inode->i_gid = 0;
-		inode->i_mode = table->mode;
-		if (table->proc_handler) {
-			inode->i_mode |= S_IFREG;
-			inode->i_nlink = 1;
-		} else {
-			inode->i_mode |= S_IFDIR;
-			inode->i_nlink = 0;	/* It is too hard to figure out */
-		}
-	}
-}
-
-static struct inode *proc_sys_make_inode(struct inode *dir, struct ctl_table *table)
+static struct inode *proc_sys_make_inode(struct super_block *sb,
+		struct ctl_table_header *head, struct ctl_table *table)
 {
 	struct inode *inode;
-	struct proc_inode *dir_ei, *ei;
-	int depth;
+	struct proc_inode *ei;
 
-	inode = new_inode(dir->i_sb);
+	inode = new_inode(sb);
 	if (!inode)
 		goto out;
 
-	/* A directory is always one deeper than it's parent */
-	dir_ei = PROC_I(dir);
-	depth = dir_ei->fd + 1;
-
+	sysctl_head_get(head);
 	ei = PROC_I(inode);
-	ei->fd = depth;
+	ei->sysctl = head;
+	ei->sysctl_entry = table;
+
 	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
-	inode->i_op = &proc_sys_inode_operations;
-	inode->i_fop = &proc_sys_file_operations;
 	inode->i_flags |= S_PRIVATE; /* tell selinux to ignore this inode */
-	proc_sys_refresh_inode(inode, table);
+	inode->i_mode = table->mode;
+	if (!table->child) {
+		inode->i_mode |= S_IFREG;
+		inode->i_op = &proc_sys_inode_operations;
+		inode->i_fop = &proc_sys_file_operations;
+	} else {
+		inode->i_mode |= S_IFDIR;
+		inode->i_nlink = 0;
+		inode->i_op = &proc_sys_dir_operations;
+		inode->i_fop = &proc_sys_dir_file_operations;
+	}
 out:
 	return inode;
 }
 
-static struct dentry *proc_sys_ancestor(struct dentry *dentry, int depth)
-{
-	for (;;) {
-		struct proc_inode *ei;
-
-		ei = PROC_I(dentry->d_inode);
-		if (ei->fd == depth)
-			break; /* found */
-
-		dentry = dentry->d_parent;
-	}
-	return dentry;
-}
-
-static struct ctl_table *proc_sys_lookup_table_one(struct ctl_table *table,
-							struct qstr *name)
+static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
 {
 	int len;
-	for ( ; table->ctl_name || table->procname; table++) {
+	for ( ; p->ctl_name || p->procname; p++) {
 
-		if (!table->procname)
+		if (!p->procname)
 			continue;
 
-		len = strlen(table->procname);
+		len = strlen(p->procname);
 		if (len != name->len)
 			continue;
 
-		if (memcmp(table->procname, name->name, len) != 0)
+		if (memcmp(p->procname, name->name, len) != 0)
 			continue;
 
 		/* I have a match */
-		return table;
+		return p;
 	}
 	return NULL;
 }
 
-static struct ctl_table *proc_sys_lookup_table(struct dentry *dentry,
-						struct ctl_table *table)
+struct ctl_table_header *grab_header(struct inode *inode)
 {
-	struct dentry *ancestor;
-	struct proc_inode *ei;
-	int depth, i;
+	if (PROC_I(inode)->sysctl)
+		return sysctl_head_grab(PROC_I(inode)->sysctl);
+	else
+		return sysctl_head_next(NULL);
+}
 
-	ei = PROC_I(dentry->d_inode);
-	depth = ei->fd;
+static struct dentry *proc_sys_lookup(struct inode *dir, struct dentry *dentry,
+					struct nameidata *nd)
+{
+	struct ctl_table_header *head = grab_header(dir);
+	struct ctl_table *table = PROC_I(dir)->sysctl_entry;
+	struct ctl_table_header *h = NULL;
+	struct qstr *name = &dentry->d_name;
+	struct ctl_table *p;
+	struct inode *inode;
+	struct dentry *err = ERR_PTR(-ENOENT);
 
-	if (depth == 0)
-		return table;
+	if (IS_ERR(head))
+		return ERR_CAST(head);
 
-	for (i = 1; table && (i <= depth); i++) {
-		ancestor = proc_sys_ancestor(dentry, i);
-		table = proc_sys_lookup_table_one(table, &ancestor->d_name);
-		if (table)
-			table = table->child;
+	if (table && !table->child) {
+		WARN_ON(1);
+		goto out;
 	}
-	return table;
-
-}
-static struct ctl_table *proc_sys_lookup_entry(struct dentry *dparent,
-						struct qstr *name,
-						struct ctl_table *table)
-{
-	table = proc_sys_lookup_table(dparent, table);
-	if (table)
-		table = proc_sys_lookup_table_one(table, name);
-	return table;
-}
 
-static struct ctl_table *do_proc_sys_lookup(struct dentry *parent,
-						struct qstr *name,
-						struct ctl_table_header **ptr)
-{
-	struct ctl_table_header *head;
-	struct ctl_table *table = NULL;
+	table = table ? table->child : head->ctl_table;
 
-	for (head = sysctl_head_next(NULL); head;
-			head = sysctl_head_next(head)) {
-		table = proc_sys_lookup_entry(parent, name, head->ctl_table);
-		if (table)
-			break;
+	p = find_in_table(table, name);
+	if (!p) {
+		for (h = sysctl_head_next(NULL); h; h = sysctl_head_next(h)) {
+			if (h->attached_to != table)
+				continue;
+			p = find_in_table(h->attached_by, name);
+			if (p)
+				break;
+		}
 	}
-	*ptr = head;
-	return table;
-}
-
-static struct dentry *proc_sys_lookup(struct inode *dir, struct dentry *dentry,
-					struct nameidata *nd)
-{
-	struct ctl_table_header *head;
-	struct inode *inode;
-	struct dentry *err;
-	struct ctl_table *table;
 
-	err = ERR_PTR(-ENOENT);
-	table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head);
-	if (!table)
+	if (!p)
 		goto out;
 
 	err = ERR_PTR(-ENOMEM);
-	inode = proc_sys_make_inode(dir, table);
+	inode = proc_sys_make_inode(dir->i_sb, h ? h : head, p);
+	if (h)
+		sysctl_head_finish(h);
+
 	if (!inode)
 		goto out;
 
@@ -168,22 +129,14 @@ out:
 static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf,
 		size_t count, loff_t *ppos, int write)
 {
-	struct dentry *dentry = filp->f_dentry;
-	struct ctl_table_header *head;
-	struct ctl_table *table;
+	struct inode *inode = filp->f_path.dentry->d_inode;
+	struct ctl_table_header *head = grab_header(inode);
+	struct ctl_table *table = PROC_I(inode)->sysctl_entry;
 	ssize_t error;
 	size_t res;
 
-	table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head);
-	/* Has the sysctl entry disappeared on us? */
-	error = -ENOENT;
-	if (!table)
-		goto out;
-
-	/* Has the sysctl entry been replaced by a directory? */
-	error = -EISDIR;
-	if (!table->proc_handler)
-		goto out;
+	if (IS_ERR(head))
+		return PTR_ERR(head);
 
 	/*
 	 * At this point we know that the sysctl was not unregistered
@@ -193,6 +146,11 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf,
 	if (sysctl_perm(head->root, table, write ? MAY_WRITE : MAY_READ))
 		goto out;
 
+	/* if that can happen at all, it should be -EINVAL, not -EISDIR */
+	error = -EINVAL;
+	if (!table->proc_handler)
+		goto out;
+
 	/* careful: calling conventions are nasty here */
 	res = count;
 	error = table->proc_handler(table, write, filp, buf, &res, ppos);
@@ -218,82 +176,86 @@ static ssize_t proc_sys_write(struct file *filp, const char __user *buf,
 
 
 static int proc_sys_fill_cache(struct file *filp, void *dirent,
-				filldir_t filldir, struct ctl_table *table)
+				filldir_t filldir,
+				struct ctl_table_header *head,
+				struct ctl_table *table)
 {
-	struct ctl_table_header *head;
-	struct ctl_table *child_table = NULL;
 	struct dentry *child, *dir = filp->f_path.dentry;
 	struct inode *inode;
 	struct qstr qname;
 	ino_t ino = 0;
 	unsigned type = DT_UNKNOWN;
-	int ret;
 
 	qname.name = table->procname;
 	qname.len  = strlen(table->procname);
 	qname.hash = full_name_hash(qname.name, qname.len);
 
-	/* Suppress duplicates.
-	 * Only fill a directory entry if it is the value that
-	 * an ordinary lookup of that name returns.  Hide all
-	 * others.
-	 *
-	 * If we ever cache this translation in the dcache
-	 * I should do a dcache lookup first.  But for now
-	 * it is just simpler not to.
-	 */
-	ret = 0;
-	child_table = do_proc_sys_lookup(dir, &qname, &head);
-	sysctl_head_finish(head);
-	if (child_table != table)
-		return 0;
-
 	child = d_lookup(dir, &qname);
 	if (!child) {
-		struct dentry *new;
-		new = d_alloc(dir, &qname);
-		if (new) {
-			inode = proc_sys_make_inode(dir->d_inode, table);
-			if (!inode)
-				child = ERR_PTR(-ENOMEM);
-			else {
-				new->d_op = &proc_sys_dentry_operations;
-				d_add(new, inode);
+		child = d_alloc(dir, &qname);
+		if (child) {
+			inode = proc_sys_make_inode(dir->d_sb, head, table);
+			if (!inode) {
+				dput(child);
+				return -ENOMEM;
+			} else {
+				child->d_op = &proc_sys_dentry_operations;
+				d_add(child, inode);
 			}
-			if (child)
-				dput(new);
-			else
-				child = new;
+		} else {
+			return -ENOMEM;
 		}
 	}
-	if (!child || IS_ERR(child) || !child->d_inode)
-		goto end_instantiate;
 	inode = child->d_inode;
-	if (inode) {
-		ino  = inode->i_ino;
-		type = inode->i_mode >> 12;
-	}
+	ino  = inode->i_ino;
+	type = inode->i_mode >> 12;
 	dput(child);
-end_instantiate:
-	if (!ino)
-		ino= find_inode_number(dir, &qname);
-	if (!ino)
-		ino = 1;
-	return filldir(dirent, qname.name, qname.len, filp->f_pos, ino, type);
+	return !!filldir(dirent, qname.name, qname.len, filp->f_pos, ino, type);
+}
+
+static int scan(struct ctl_table_header *head, ctl_table *table,
+		unsigned long *pos, struct file *file,
+		void *dirent, filldir_t filldir)
+{
+
+	for (; table->ctl_name || table->procname; table++, (*pos)++) {
+		int res;
+
+		/* Can't do anything without a proc name */
+		if (!table->procname)
+			continue;
+
+		if (*pos < file->f_pos)
+			continue;
+
+		res = proc_sys_fill_cache(file, dirent, filldir, head, table);
+		if (res)
+			return res;
+
+		file->f_pos = *pos + 1;
+	}
+	return 0;
 }
 
 static int proc_sys_readdir(struct file *filp, void *dirent, filldir_t filldir)
 {
-	struct dentry *dentry = filp->f_dentry;
+	struct dentry *dentry = filp->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
-	struct ctl_table_header *head = NULL;
-	struct ctl_table *table;
+	struct ctl_table_header *head = grab_header(inode);
+	struct ctl_table *table = PROC_I(inode)->sysctl_entry;
+	struct ctl_table_header *h = NULL;
 	unsigned long pos;
-	int ret;
+	int ret = -EINVAL;
+
+	if (IS_ERR(head))
+		return PTR_ERR(head);
 
-	ret = -ENOTDIR;
-	if (!S_ISDIR(inode->i_mode))
+	if (table && !table->child) {
+		WARN_ON(1);
 		goto out;
+	}
+
+	table = table ? table->child : head->ctl_table;
 
 	ret = 0;
 	/* Avoid a switch here: arm builds fail with missing __cmpdi2 */
@@ -311,30 +273,17 @@ static int proc_sys_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	}
 	pos = 2;
 
-	/* - Find each instance of the directory
-	 * - Read all entries in each instance
-	 * - Before returning an entry to user space lookup the entry
-	 *   by name and if I find a different entry don't return
-	 *   this one because it means it is a buried dup.
-	 * For sysctl this should only happen for directory entries.
-	 */
-	for (head = sysctl_head_next(NULL); head; head = sysctl_head_next(head)) {
-		table = proc_sys_lookup_table(dentry, head->ctl_table);
+	ret = scan(head, table, &pos, filp, dirent, filldir);
+	if (ret)
+		goto out;
 
-		if (!table)
+	for (h = sysctl_head_next(NULL); h; h = sysctl_head_next(h)) {
+		if (h->attached_to != table)
 			continue;
-
-		for (; table->ctl_name || table->procname; table++, pos++) {
-			/* Can't do anything without a proc name */
-			if (!table->procname)
-				continue;
-
-			if (pos < filp->f_pos)
-				continue;
-
-			if (proc_sys_fill_cache(filp, dirent, filldir, table) < 0)
-				goto out;
-			filp->f_pos = pos + 1;
+		ret = scan(h, h->attached_by, &pos, filp, dirent, filldir);
+		if (ret) {
+			sysctl_head_finish(h);
+			break;
 		}
 	}
 	ret = 1;
@@ -349,47 +298,18 @@ static int proc_sys_permission(struct inode *inode, int mask, struct nameidata *
 	 * sysctl entries that are not writeable,
 	 * are _NOT_ writeable, capabilities or not.
 	 */
-	struct ctl_table_header *head;
-	struct ctl_table *table;
-	struct dentry *dentry;
-	int mode;
-	int depth;
+	struct ctl_table_header *head = grab_header(inode);
+	struct ctl_table *table = PROC_I(inode)->sysctl_entry;
 	int error;
 
-	head = NULL;
-	depth = PROC_I(inode)->fd;
-
-	/* First check the cached permissions, in case we don't have
-	 * enough information to lookup the sysctl table entry.
-	 */
-	error = -EACCES;
-	mode = inode->i_mode;
-
-	if (current->euid == 0)
-		mode >>= 6;
-	else if (in_group_p(0))
-		mode >>= 3;
-
-	if ((mode & mask & (MAY_READ|MAY_WRITE|MAY_EXEC)) == mask)
-		error = 0;
-
-	/* If we can't get a sysctl table entry the permission
-	 * checks on the cached mode will have to be enough.
-	 */
-	if (!nd || !depth)
-		goto out;
+	if (IS_ERR(head))
+		return PTR_ERR(head);
 
-	dentry = nd->path.dentry;
-	table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head);
+	if (!table) /* global root - r-xr-xr-x */
+		error = mask & MAY_WRITE ? -EACCES : 0;
+	else /* Use the permissions on the sysctl table entry */
+		error = sysctl_perm(head->root, table, mask);
 
-	/* If the entry does not exist deny permission */
-	error = -EACCES;
-	if (!table)
-		goto out;
-
-	/* Use the permissions on the sysctl table entry */
-	error = sysctl_perm(head->root, table, mask);
-out:
 	sysctl_head_finish(head);
 	return error;
 }
@@ -409,33 +329,70 @@ static int proc_sys_setattr(struct dentry *dentry, struct iattr *attr)
 	return error;
 }
 
-/* I'm lazy and don't distinguish between files and directories,
- * until access time.
- */
+static int proc_sys_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
+{
+	struct inode *inode = dentry->d_inode;
+	struct ctl_table_header *head = grab_header(inode);
+	struct ctl_table *table = PROC_I(inode)->sysctl_entry;
+
+	if (IS_ERR(head))
+		return PTR_ERR(head);
+
+	generic_fillattr(inode, stat);
+	if (table)
+		stat->mode = (stat->mode & S_IFMT) | table->mode;
+
+	sysctl_head_finish(head);
+	return 0;
+}
+
 static const struct file_operations proc_sys_file_operations = {
 	.read		= proc_sys_read,
 	.write		= proc_sys_write,
+};
+
+static const struct file_operations proc_sys_dir_file_operations = {
 	.readdir	= proc_sys_readdir,
 };
 
 static const struct inode_operations proc_sys_inode_operations = {
+	.permission	= proc_sys_permission,
+	.setattr	= proc_sys_setattr,
+	.getattr	= proc_sys_getattr,
+};
+
+static const struct inode_operations proc_sys_dir_operations = {
 	.lookup		= proc_sys_lookup,
 	.permission	= proc_sys_permission,
 	.setattr	= proc_sys_setattr,
+	.getattr	= proc_sys_getattr,
 };
 
 static int proc_sys_revalidate(struct dentry *dentry, struct nameidata *nd)
 {
-	struct ctl_table_header *head;
-	struct ctl_table *table;
-	table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head);
-	proc_sys_refresh_inode(dentry->d_inode, table);
-	sysctl_head_finish(head);
-	return !!table;
+	return !PROC_I(dentry->d_inode)->sysctl->unregistering;
+}
+
+static int proc_sys_delete(struct dentry *dentry)
+{
+	return !!PROC_I(dentry->d_inode)->sysctl->unregistering;
+}
+
+static int proc_sys_compare(struct dentry *dir, struct qstr *qstr,
+			    struct qstr *name)
+{
+	struct dentry *dentry = container_of(qstr, struct dentry, d_name);
+	if (qstr->len != name->len)
+		return 1;
+	if (memcmp(qstr->name, name->name, name->len))
+		return 1;
+	return !sysctl_is_seen(PROC_I(dentry->d_inode)->sysctl);
 }
 
 static struct dentry_operations proc_sys_dentry_operations = {
 	.d_revalidate	= proc_sys_revalidate,
+	.d_delete	= proc_sys_delete,
+	.d_compare	= proc_sys_compare,
 };
 
 static struct proc_dir_entry *proc_sys_root;
@@ -443,8 +400,8 @@ static struct proc_dir_entry *proc_sys_root;
 int proc_sys_init(void)
 {
 	proc_sys_root = proc_mkdir("sys", NULL);
-	proc_sys_root->proc_iops = &proc_sys_inode_operations;
-	proc_sys_root->proc_fops = &proc_sys_file_operations;
+	proc_sys_root->proc_iops = &proc_sys_dir_operations;
+	proc_sys_root->proc_fops = &proc_sys_dir_file_operations;
 	proc_sys_root->nlink = 0;
 	return 0;
 }
diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h
index f560d1705afe..fb61850d1cfc 100644
--- a/include/linux/proc_fs.h
+++ b/include/linux/proc_fs.h
@@ -282,11 +282,16 @@ union proc_op {
 		struct task_struct *task);
 };
 
+struct ctl_table_header;
+struct ctl_table;
+
 struct proc_inode {
 	struct pid *pid;
 	int fd;
 	union proc_op op;
 	struct proc_dir_entry *pde;
+	struct ctl_table_header *sysctl;
+	struct ctl_table *sysctl_entry;
 	struct inode vfs_inode;
 };
 
diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h
index 3f6599aeb0db..d0437f36921f 100644
--- a/include/linux/sysctl.h
+++ b/include/linux/sysctl.h
@@ -961,6 +961,7 @@ struct ctl_table_header;
 
 extern void sysctl_head_get(struct ctl_table_header *);
 extern void sysctl_head_put(struct ctl_table_header *);
+extern int sysctl_is_seen(struct ctl_table_header *);
 extern struct ctl_table_header *sysctl_head_grab(struct ctl_table_header *);
 extern struct ctl_table_header *sysctl_head_next(struct ctl_table_header *prev);
 extern struct ctl_table_header *__sysctl_head_next(struct nsproxy *namespaces,
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index c9a0af887033..ff5abcca5ddf 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -1932,6 +1932,21 @@ void unregister_sysctl_table(struct ctl_table_header * header)
 	spin_unlock(&sysctl_lock);
 }
 
+int sysctl_is_seen(struct ctl_table_header *p)
+{
+	struct ctl_table_set *set = p->set;
+	int res;
+	spin_lock(&sysctl_lock);
+	if (p->unregistering)
+		res = 0;
+	else if (!set->is_seen)
+		res = 1;
+	else
+		res = set->is_seen(set);
+	spin_unlock(&sysctl_lock);
+	return res;
+}
+
 void setup_sysctl_set(struct ctl_table_set *p,
 	struct ctl_table_set *parent,
 	int (*is_seen)(struct ctl_table_set *))
-- 
cgit v1.2.3