fs: remove f_version

Now that detecting concurrent seeks is done by the filesystems that
require it we can remove f_version and free up 8 bytes for future
extensions.

Link: https://lore.kernel.org/r/[e-mail]
Reviewed-by: Jan Kara <[e-mail]>
Reviewed-by: Jeff Layton <[e-mail]>
Signed-off-by: Christian Brauner <[e-mail]>

fs/read_write.c

···
 
 	if (offset != file->f_pos) {
 		file->f_pos = offset;
-		*cookie = 0;
+		if (cookie)
+			*cookie = 0;
 	}
 	return offset;
 }
···
  */
 loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize)
 {
-	return vfs_setpos_cookie(file, offset, maxsize, &file->f_version);
+	return vfs_setpos_cookie(file, offset, maxsize, NULL);
 }
 EXPORT_SYMBOL(vfs_setpos);
 
···
 	}
 	retval = -EINVAL;
 	if (offset >= 0 || unsigned_offsets(file)) {
-		if (offset != file->f_pos) {
+		if (offset != file->f_pos)
 			file->f_pos = offset;
-			file->f_version = 0;
-		}
 		retval = offset;
 	}
 out:

include/linux/fs.h

···
  * @f_pos_lock: lock protecting file position
  * @f_pipe: specific to pipes
  * @f_pos: file position
- * @f_version: file version
  * @f_security: LSM security context of this file
  * @f_owner: file owner
  * @f_wb_err: writeback error
···
 		u64			f_pipe;
 	};
 	loff_t				f_pos;
-	u64				f_version;
-	/* --- cacheline 2 boundary (128 bytes) --- */
 #ifdef CONFIG_SECURITY
 	void				*f_security;
 #endif
+	/* --- cacheline 2 boundary (128 bytes) --- */
 	struct fown_struct		*f_owner;
 	errseq_t			f_wb_err;
 	errseq_t			f_sb_err;