From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cn.fujitsu.com ([222.73.24.84]:65097 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1754336Ab2LLEO4 (ORCPT ); Tue, 11 Dec 2012 23:14:56 -0500 Message-ID: <50C804D9.5060700@cn.fujitsu.com> Date: Wed, 12 Dec 2012 12:15:21 +0800 From: Miao Xie Reply-To: miaox@cn.fujitsu.com MIME-Version: 1.0 To: Jeff Liu CC: linux-btrfs@vger.kernel.org, anand.jain@oracle.com Subject: Re: [RESEND PATCH V4 2/2] Btrfs: Add a new ioctl to change the label of a mounted filesystem In-Reply-To: <50C7F894.9000709@oracle.com> Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-btrfs-owner@vger.kernel.org List-ID: References: <50C7F894.9000709@oracle.com> On wed, 12 Dec 2012 11:23:00 +0800, Jeff Liu wrote: > Add a new ioctl BTRFS_FS_SETLABEL to change the label of a mounted filesystem. > > > Signed-off-by: Jie Liu > Signed-off-by: Anand Jain > > --- > fs/btrfs/ioctl.c | 34 ++++++++++++++++++++++++++++++++++ > fs/btrfs/ioctl.h | 2 ++ > 2 files changed, 36 insertions(+) > > diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c > index b0a5e17..ebbf634 100644 > --- a/fs/btrfs/ioctl.c > +++ b/fs/btrfs/ioctl.c > @@ -3711,6 +3711,38 @@ static int btrfs_ioctl_get_fslabel(struct btrfs_root *root, void __user *arg) > return ret ? -EFAULT : 0; > } > > +static int btrfs_ioctl_set_fslabel(struct btrfs_root *root, void __user *arg) > +{ > + struct btrfs_super_block *super_block = root->fs_info->super_copy; > + struct btrfs_trans_handle *trans; > + char label[BTRFS_LABEL_SIZE]; > + int ret = 0; > + > + if (!capable(CAP_SYS_ADMIN)) > + return -EPERM; > + > + if (btrfs_root_readonly(root)) > + return -EROFS; Since this label is a global parameter, we needn't check the root is R/O or not. The check we need is the write access of the fs, so we need call mnt_want_write_file() > + > + if (copy_from_user(label, arg, sizeof(label))) > + return -EFAULT; > + > + label[BTRFS_LABEL_SIZE - 1] = '\0'; I don't think setting the last byte to '\0' is a good way, because the user would be very strange that the new label is truncated, but the program told him that it was done successfully. > + > + mutex_lock(&root->fs_info->volume_mutex); > + trans = btrfs_start_transaction(root, 1); > + if (IS_ERR(trans)) { > + ret = PTR_ERR(trans); > + goto out_unlock; > + } > + strcpy(super_block->label, label); > + btrfs_end_transaction(trans, root); > + > +out_unlock: > + mutex_unlock(&root->fs_info->volume_mutex); > + return ret; > +} > + > long btrfs_ioctl(struct file *file, unsigned int > cmd, unsigned long arg) > { > @@ -3811,6 +3843,8 @@ long btrfs_ioctl(struct file *file, unsigned int > return btrfs_ioctl_qgroup_limit(root, argp); > case BTRFS_IOC_GET_FSLABEL: > return btrfs_ioctl_get_fslabel(root, argp); > + case BTRFS_IOC_SET_FSLABEL: > + return btrfs_ioctl_set_fslabel(root, argp); > } > > return -ENOTTY; > diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h > index 98f896f..cc4e657 100644 > --- a/fs/btrfs/ioctl.h > +++ b/fs/btrfs/ioctl.h > @@ -455,4 +455,6 @@ struct btrfs_ioctl_send_args { > struct btrfs_ioctl_get_dev_stats) > #define BTRFS_IOC_GET_FSLABEL _IOR(BTRFS_IOCTL_MAGIC, 53, \ > char[BTRFS_LABEL_SIZE]) > +#define BTRFS_IOC_SET_FSLABEL _IOW(BTRFS_IOCTL_MAGIC, 54, \ > + char[BTRFS_LABEL_SIZE]) > #endif number 50 has been reserved for this feature. Thanks Miao