udf: Replace bkl with the UDF_I(inode)->i_data_sem for protect udf_in…

…ode_info struct

Replace bkl with the UDF_I(inode)->i_data_sem rw semaphore in
udf_release_file(), udf_symlink(), udf_symlink_filler(), udf_get_block(),
udf_block_map(), and udf_setattr(). The rule now is that any operation
on regular file's or symlink's extents (or generally allocation information
including goal block) needs to hold i_data_sem.

This work was supported by a hardware donation from the CE Linux Forum.

Signed-off-by: Alessio Igor Bogani <abogani@texware.it>
Signed-off-by: Jan Kara <jack@suse.cz>

fs/udf/file.c

@@ -204,10 +204,10 @@ static int udf_release_file(struct inode *inode, struct file *filp)
 {
 	if (filp->f_mode & FMODE_WRITE) {
 		mutex_lock(&inode->i_mutex);
-		lock_kernel();
+		down_write(&UDF_I(inode)->i_data_sem);
 		udf_discard_prealloc(inode);
 		udf_truncate_tail_extent(inode);
-		unlock_kernel();
+		up_write(&UDF_I(inode)->i_data_sem);
 		mutex_unlock(&inode->i_mutex);
 	}
 	return 0;

fs/udf/inode.c

@@ -301,10 +301,9 @@ static int udf_get_block(struct inode *inode, sector_t block,
 	err = -EIO;
 	new = 0;
 	bh = NULL;
-
-	lock_kernel();
-
 	iinfo = UDF_I(inode);
+
+	down_write(&iinfo->i_data_sem);
 	if (block == iinfo->i_next_alloc_block + 1) {
 		iinfo->i_next_alloc_block++;
 		iinfo->i_next_alloc_goal++;
@@ -323,7 +322,7 @@ static int udf_get_block(struct inode *inode, sector_t block,
 	map_bh(bh_result, inode->i_sb, phys);
 
 abort:
-	unlock_kernel();
+	up_write(&iinfo->i_data_sem);
 	return err;
 }
 
@@ -1021,16 +1020,16 @@ void udf_truncate(struct inode *inode)
 	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
 		return;
 
-	lock_kernel();
 	iinfo = UDF_I(inode);
 	if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) {
+		down_write(&iinfo->i_data_sem);
 		if (inode->i_sb->s_blocksize <
 				(udf_file_entry_alloc_offset(inode) +
 				 inode->i_size)) {
 			udf_expand_file_adinicb(inode, inode->i_size, &err);
 			if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) {
 				inode->i_size = iinfo->i_lenAlloc;
-				unlock_kernel();
+				up_write(&iinfo->i_data_sem);
 				return;
 			} else
 				udf_truncate_extents(inode);
@@ -1041,18 +1040,20 @@ void udf_truncate(struct inode *inode)
 				offset - udf_file_entry_alloc_offset(inode));
 			iinfo->i_lenAlloc = inode->i_size;
 		}
+		up_write(&iinfo->i_data_sem);
 	} else {
 		block_truncate_page(inode->i_mapping, inode->i_size,
 				    udf_get_block);
+		down_write(&iinfo->i_data_sem);
 		udf_truncate_extents(inode);
+		up_write(&iinfo->i_data_sem);
 	}
 
 	inode->i_mtime = inode->i_ctime = current_fs_time(inode->i_sb);
 	if (IS_SYNC(inode))
 		udf_sync_inode(inode);
 	else
 		mark_inode_dirty(inode);
-	unlock_kernel();
 }
 
 static void __udf_read_inode(struct inode *inode)
@@ -2043,15 +2044,15 @@ long udf_block_map(struct inode *inode, sector_t block)
 	struct extent_position epos = {};
 	int ret;
 
-	lock_kernel();
+	down_read(&UDF_I(inode)->i_data_sem);
 
 	if (inode_bmap(inode, block, &epos, &eloc, &elen, &offset) ==
 						(EXT_RECORDED_ALLOCATED >> 30))
 		ret = udf_get_lb_pblock(inode->i_sb, &eloc, offset);
 	else
 		ret = 0;
 
-	unlock_kernel();
+	up_read(&UDF_I(inode)->i_data_sem);
 	brelse(epos.bh);
 
 	if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_VARCONV))

fs/udf/namei.c

@@ -893,18 +893,18 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry,
 	struct udf_inode_info *iinfo;
 	struct super_block *sb = dir->i_sb;
 
-	lock_kernel();
 	inode = udf_new_inode(dir, S_IFLNK | S_IRWXUGO, &err);
 	if (!inode)
 		goto out;
 
+	iinfo = UDF_I(inode);
+	down_write(&iinfo->i_data_sem);
 	name = kmalloc(UDF_NAME_LEN, GFP_NOFS);
 	if (!name) {
 		err = -ENOMEM;
 		goto out_no_entry;
 	}
 
-	iinfo = UDF_I(inode);
 	inode->i_data.a_ops = &udf_symlink_aops;
 	inode->i_op = &udf_symlink_inode_operations;
 
@@ -1024,6 +1024,7 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry,
 	udf_write_fi(dir, &cfi, fi, &fibh, NULL, NULL);
 	if (UDF_I(dir)->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB)
 		mark_inode_dirty(dir);
+	up_write(&iinfo->i_data_sem);
 	if (fibh.sbh != fibh.ebh)
 		brelse(fibh.ebh);
 	brelse(fibh.sbh);
@@ -1032,10 +1033,10 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry,
 
 out:
 	kfree(name);
-	unlock_kernel();
 	return err;
 
 out_no_entry:
+	up_write(&iinfo->i_data_sem);
 	inode_dec_link_count(inode);
 	iput(inode);
 	goto out;

fs/udf/super.c

@@ -135,6 +135,7 @@ static struct inode *udf_alloc_inode(struct super_block *sb)
 	ei->i_next_alloc_block = 0;
 	ei->i_next_alloc_goal = 0;
 	ei->i_strat4096 = 0;
+	init_rwsem(&ei->i_data_sem);
 
 	return &ei->vfs_inode;
 }

fs/udf/symlink.c

@@ -27,7 +27,6 @@
 #include <linux/mm.h>
 #include <linux/stat.h>
 #include <linux/pagemap.h>
-#include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
 #include "udf_i.h"
 
@@ -78,13 +77,16 @@ static int udf_symlink_filler(struct file *file, struct page *page)
 	int err = -EIO;
 	unsigned char *p = kmap(page);
 	struct udf_inode_info *iinfo;
+	uint32_t pos;
 
-	lock_kernel();
 	iinfo = UDF_I(inode);
+	pos = udf_block_map(inode, 0);
+
+	down_read(&iinfo->i_data_sem);
 	if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) {
 		symlink = iinfo->i_ext.i_data + iinfo->i_lenEAttr;
 	} else {
-		bh = sb_bread(inode->i_sb, udf_block_map(inode, 0));
+		bh = sb_bread(inode->i_sb, pos);
 
 		if (!bh)
 			goto out;
@@ -95,14 +97,14 @@ static int udf_symlink_filler(struct file *file, struct page *page)
 	udf_pc_to_char(inode->i_sb, symlink, inode->i_size, p);
 	brelse(bh);
 
-	unlock_kernel();
+	up_read(&iinfo->i_data_sem);
 	SetPageUptodate(page);
 	kunmap(page);
 	unlock_page(page);
 	return 0;
 
 out:
-	unlock_kernel();
+	up_read(&iinfo->i_data_sem);
 	SetPageError(page);
 	kunmap(page);
 	unlock_page(page);

fs/udf/udf_i.h

@@ -1,6 +1,18 @@
 #ifndef _UDF_I_H
 #define _UDF_I_H
 
+/*
+ * The i_data_sem and i_mutex serve for protection of allocation information
+ * of a regular files and symlinks. This includes all extents belonging to
+ * the file/symlink, a fact whether data are in-inode or in external data
+ * blocks, preallocation, goal block information... When extents are read,
+ * i_mutex or i_data_sem must be held (for reading is enough in case of
+ * i_data_sem). When extents are changed, i_data_sem must be held for writing
+ * and also i_mutex must be held.
+ *
+ * For directories i_mutex is used for all the necessary protection.
+ */
+
 struct udf_inode_info {
 	struct timespec		i_crtime;
 	/* Physical address of inode */
@@ -21,6 +33,7 @@ struct udf_inode_info {
 		struct long_ad		*i_lad;
 		__u8		*i_data;
 	} i_ext;
+	struct rw_semaphore	i_data_sem;
 	struct inode vfs_inode;
 };