[PATCH] Don't hold i_sem on swapfiles

We permanently hold the i_sem of swapfiles so that nobody can addidentally
ftruncate them, causing subsequent filesystem destruction.

Problem is, it's fairly easy for things like backup applications to get
stuck onthe swapfile, sleeping until someone does a swapoff.

So take all that out again and add a new S_SWAPFILE inode flag.  Test that
in the truncate path and refuse to truncate an in-use swapfile.

Synchronisation between swapon and truncate is via i_sem.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

include/linux/fs.h

@@ -140,6 +140,7 @@ extern int leases_enable, dir_notify_enable, lease_break_time;
 #define S_NOQUOTA	64	/* Inode is not counted to quota */
 #define S_DIRSYNC	128	/* Directory modifications are synchronous */
 #define S_NOCMTIME	256	/* Do not update file c/mtime */
+#define S_SWAPFILE	512	/* Do not truncate: swapon got its bmaps */
 
 /*
  * Note that nosuid etc flags are inode-specific: setting some file-system
@@ -174,6 +175,7 @@ extern int leases_enable, dir_notify_enable, lease_break_time;
 
 #define IS_DEADDIR(inode)	((inode)->i_flags & S_DEAD)
 #define IS_NOCMTIME(inode)	((inode)->i_flags & S_NOCMTIME)
+#define IS_SWAPFILE(inode)	((inode)->i_flags & S_SWAPFILE)
 
 /* the read-only stuff doesn't really belong here, but any other place is
    probably as bad and I don't want to create yet another include file. */

mm/memory.c

@@ -1223,6 +1223,12 @@ int vmtruncate(struct inode * inode, loff_t offset)
 
 	if (inode->i_size < offset)
 		goto do_expand;
+	/*
+	 * truncation of in-use swapfiles is disallowed - it would cause
+	 * subsequent swapout to scribble on the now-freed blocks.
+	 */
+	if (IS_SWAPFILE(inode))
+		goto out_busy;
 	i_size_write(inode, offset);
 	unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1);
 	truncate_inode_pages(mapping, offset);
@@ -1233,7 +1239,7 @@ int vmtruncate(struct inode * inode, loff_t offset)
 	if (limit != RLIM_INFINITY && offset > limit)
 		goto out_sig;
 	if (offset > inode->i_sb->s_maxbytes)
-		goto out;
+		goto out_big;
 	i_size_write(inode, offset);
 
 out_truncate:
@@ -1242,8 +1248,10 @@ int vmtruncate(struct inode * inode, loff_t offset)
 	return 0;
 out_sig:
 	send_sig(SIGXFSZ, current, 0);
-out:
+out_big:
 	return -EFBIG;
+out_busy:
+	return -ETXTBSY;
 }
 
 EXPORT_SYMBOL(vmtruncate);

mm/swapfile.c

@@ -1072,6 +1072,7 @@ asmlinkage long sys_swapoff(const char __user * specialfile)
 	unsigned short *swap_map;
 	struct file *swap_file, *victim;
 	struct address_space *mapping;
+	struct inode *inode;
 	char * pathname;
 	int i, type, prev;
 	int err;
@@ -1165,12 +1166,15 @@ asmlinkage long sys_swapoff(const char __user * specialfile)
 	swap_list_unlock();
 	up(&swapon_sem);
 	vfree(swap_map);
-	if (S_ISBLK(mapping->host->i_mode)) {
-		struct block_device *bdev = I_BDEV(mapping->host);
+	inode = mapping->host;
+	if (S_ISBLK(inode->i_mode)) {
+		struct block_device *bdev = I_BDEV(inode);
 		set_blocksize(bdev, p->old_block_size);
 		bd_release(bdev);
 	} else {
-		up(&mapping->host->i_sem);
+		down(&inode->i_sem);
+		inode->i_flags &= ~S_SWAPFILE;
+		up(&inode->i_sem);
 	}
 	filp_close(swap_file, NULL);
 	err = 0;
@@ -1388,6 +1392,10 @@ asmlinkage long sys_swapon(const char __user * specialfile, int swap_flags)
 		p->bdev = inode->i_sb->s_bdev;
 		down(&inode->i_sem);
 		did_down = 1;
+		if (IS_SWAPFILE(inode)) {
+			error = -EBUSY;
+			goto bad_swap;
+		}
 	} else {
 		goto bad_swap;
 	}
@@ -1560,8 +1568,11 @@ asmlinkage long sys_swapon(const char __user * specialfile, int swap_flags)
 	}
 	if (name)
 		putname(name);
-	if (error && did_down)
+	if (did_down) {
+		if (!error)
+			inode->i_flags |= S_SWAPFILE;
 		up(&inode->i_sem);
+	}
 	return error;
 }