diff mbox series

[v1,01/17] btrfs: disable various operations on encrypted inodes

Message ID e7785ffe237e581a7ba7e45d2724fca4d8fa1470.1687988380.git.sweettea-kernel@dorminy.me (mailing list archive)
State Superseded
Headers show
Series btrfs: add encryption feature | expand

Commit Message

Sweet Tea Dorminy June 29, 2023, 12:35 a.m. UTC
From: Omar Sandoval <osandov@osandov.com>

Initially, only normal data extents, using the normal (non-direct) IO
path, will be encrypted. This change forbids various other bits:
- allows reflinking only if both inodes have the same encryption status
- disables direct IO on encrypted inodes
- disable inline data on encrypted inodes

Signed-off-by: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
---
 fs/btrfs/file.c    | 4 ++--
 fs/btrfs/inode.c   | 3 ++-
 fs/btrfs/reflink.c | 7 +++++++
 3 files changed, 11 insertions(+), 3 deletions(-)

Comments

Boris Burkov July 7, 2023, 11:36 p.m. UTC | #1
On Wed, Jun 28, 2023 at 08:35:24PM -0400, Sweet Tea Dorminy wrote:
> From: Omar Sandoval <osandov@osandov.com>
> 
> Initially, only normal data extents, using the normal (non-direct) IO
> path, will be encrypted. This change forbids various other bits:
> - allows reflinking only if both inodes have the same encryption status
> - disables direct IO on encrypted inodes
> - disable inline data on encrypted inodes
> 
> Signed-off-by: Omar Sandoval <osandov@osandov.com>
> Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
> ---
>  fs/btrfs/file.c    | 4 ++--
>  fs/btrfs/inode.c   | 3 ++-
>  fs/btrfs/reflink.c | 7 +++++++
>  3 files changed, 11 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
> index 392bc7d512a0..354962a7dd72 100644
> --- a/fs/btrfs/file.c
> +++ b/fs/btrfs/file.c
> @@ -1502,7 +1502,7 @@ static ssize_t btrfs_direct_write(struct kiocb *iocb, struct iov_iter *from)
>  		goto relock;
>  	}
>  
> -	if (check_direct_IO(fs_info, from, pos)) {
> +	if (IS_ENCRYPTED(inode) || check_direct_IO(fs_info, from, pos)) {
>  		btrfs_inode_unlock(BTRFS_I(inode), ilock_flags);
>  		goto buffered;
>  	}
> @@ -3741,7 +3741,7 @@ static ssize_t btrfs_direct_read(struct kiocb *iocb, struct iov_iter *to)
>  	ssize_t read = 0;
>  	ssize_t ret;
>  
> -	if (fsverity_active(inode))
> +	if (IS_ENCRYPTED(inode) || fsverity_active(inode))

What's different about fscrypt vs fsverity that makes the inode flag a
good check for encryption while verity relies on the presence of the
extra context metadata?

Is the enable model not subject to the same race where S_VERITY gets set
ahead of actually storing the verity info/item?

I think it's fine for these "skip" cases, but I imagine if we have cases
of "I want a fully ready encrypted file" the verity-style check could be
better?

>  		return 0;
>  
>  	if (check_direct_read(btrfs_sb(inode->i_sb), to, iocb->ki_pos))
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index dbbb67293e34..48eadc4f187f 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -630,7 +630,8 @@ static noinline int cow_file_range_inline(struct btrfs_inode *inode, u64 size,
>  	 * compressed) data fits in a leaf and the configured maximum inline
>  	 * size.
>  	 */
> -	if (size < i_size_read(&inode->vfs_inode) ||
> +	if (IS_ENCRYPTED(&inode->vfs_inode) ||
> +	    size < i_size_read(&inode->vfs_inode) ||
>  	    size > fs_info->sectorsize ||
>  	    data_len > BTRFS_MAX_INLINE_DATA_SIZE(fs_info) ||
>  	    data_len > fs_info->max_inline)
> diff --git a/fs/btrfs/reflink.c b/fs/btrfs/reflink.c
> index 0474bbe39da7..ad722f495c9b 100644
> --- a/fs/btrfs/reflink.c
> +++ b/fs/btrfs/reflink.c
> @@ -1,6 +1,7 @@
>  // SPDX-License-Identifier: GPL-2.0
>  
>  #include <linux/blkdev.h>
> +#include <linux/fscrypt.h>
>  #include <linux/iversion.h>
>  #include "ctree.h"
>  #include "fs.h"
> @@ -811,6 +812,12 @@ static int btrfs_remap_file_range_prep(struct file *file_in, loff_t pos_in,
>  		ASSERT(inode_in->i_sb == inode_out->i_sb);
>  	}
>  
> +	/*
> +	 * Can only reflink encrypted files if both files are encrypted.
> +	 */
> +	if (IS_ENCRYPTED(inode_in) != IS_ENCRYPTED(inode_out))
> +		return -EINVAL;
> +
>  	/* Don't make the dst file partly checksummed */
>  	if ((BTRFS_I(inode_in)->flags & BTRFS_INODE_NODATASUM) !=
>  	    (BTRFS_I(inode_out)->flags & BTRFS_INODE_NODATASUM)) {
> -- 
> 2.40.1
>
Sweet Tea Dorminy July 17, 2023, 1:42 a.m. UTC | #2
On 7/7/23 19:36, Boris Burkov wrote:
> On Wed, Jun 28, 2023 at 08:35:24PM -0400, Sweet Tea Dorminy wrote:
>> From: Omar Sandoval <osandov@osandov.com>
>>
>> Initially, only normal data extents, using the normal (non-direct) IO
>> path, will be encrypted. This change forbids various other bits:
>> - allows reflinking only if both inodes have the same encryption status
>> - disables direct IO on encrypted inodes
>> - disable inline data on encrypted inodes
>>
>> Signed-off-by: Omar Sandoval <osandov@osandov.com>
>> Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
>> ---
>>   fs/btrfs/file.c    | 4 ++--
>>   fs/btrfs/inode.c   | 3 ++-
>>   fs/btrfs/reflink.c | 7 +++++++
>>   3 files changed, 11 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
>> index 392bc7d512a0..354962a7dd72 100644
>> --- a/fs/btrfs/file.c
>> +++ b/fs/btrfs/file.c
>> @@ -1502,7 +1502,7 @@ static ssize_t btrfs_direct_write(struct kiocb *iocb, struct iov_iter *from)
>>   		goto relock;
>>   	}
>>   
>> -	if (check_direct_IO(fs_info, from, pos)) {
>> +	if (IS_ENCRYPTED(inode) || check_direct_IO(fs_info, from, pos)) {
>>   		btrfs_inode_unlock(BTRFS_I(inode), ilock_flags);
>>   		goto buffered;
>>   	}
>> @@ -3741,7 +3741,7 @@ static ssize_t btrfs_direct_read(struct kiocb *iocb, struct iov_iter *to)
>>   	ssize_t read = 0;
>>   	ssize_t ret;
>>   
>> -	if (fsverity_active(inode))
>> +	if (IS_ENCRYPTED(inode) || fsverity_active(inode))
> 
> What's different about fscrypt vs fsverity that makes the inode flag a
> good check for encryption while verity relies on the presence of the
> extra context metadata?
> 
> Is the enable model not subject to the same race where S_VERITY gets set
> ahead of actually storing the verity info/item?
> 
> I think it's fine for these "skip" cases, but I imagine if we have cases
> of "I want a fully ready encrypted file" the verity-style check could be
> better?
> 
Clear solution: enable direct encrypted IO :). I've removed this check 
altogether, thanks for the note.
diff mbox series

Patch

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 392bc7d512a0..354962a7dd72 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1502,7 +1502,7 @@  static ssize_t btrfs_direct_write(struct kiocb *iocb, struct iov_iter *from)
 		goto relock;
 	}
 
-	if (check_direct_IO(fs_info, from, pos)) {
+	if (IS_ENCRYPTED(inode) || check_direct_IO(fs_info, from, pos)) {
 		btrfs_inode_unlock(BTRFS_I(inode), ilock_flags);
 		goto buffered;
 	}
@@ -3741,7 +3741,7 @@  static ssize_t btrfs_direct_read(struct kiocb *iocb, struct iov_iter *to)
 	ssize_t read = 0;
 	ssize_t ret;
 
-	if (fsverity_active(inode))
+	if (IS_ENCRYPTED(inode) || fsverity_active(inode))
 		return 0;
 
 	if (check_direct_read(btrfs_sb(inode->i_sb), to, iocb->ki_pos))
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index dbbb67293e34..48eadc4f187f 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -630,7 +630,8 @@  static noinline int cow_file_range_inline(struct btrfs_inode *inode, u64 size,
 	 * compressed) data fits in a leaf and the configured maximum inline
 	 * size.
 	 */
-	if (size < i_size_read(&inode->vfs_inode) ||
+	if (IS_ENCRYPTED(&inode->vfs_inode) ||
+	    size < i_size_read(&inode->vfs_inode) ||
 	    size > fs_info->sectorsize ||
 	    data_len > BTRFS_MAX_INLINE_DATA_SIZE(fs_info) ||
 	    data_len > fs_info->max_inline)
diff --git a/fs/btrfs/reflink.c b/fs/btrfs/reflink.c
index 0474bbe39da7..ad722f495c9b 100644
--- a/fs/btrfs/reflink.c
+++ b/fs/btrfs/reflink.c
@@ -1,6 +1,7 @@ 
 // SPDX-License-Identifier: GPL-2.0
 
 #include <linux/blkdev.h>
+#include <linux/fscrypt.h>
 #include <linux/iversion.h>
 #include "ctree.h"
 #include "fs.h"
@@ -811,6 +812,12 @@  static int btrfs_remap_file_range_prep(struct file *file_in, loff_t pos_in,
 		ASSERT(inode_in->i_sb == inode_out->i_sb);
 	}
 
+	/*
+	 * Can only reflink encrypted files if both files are encrypted.
+	 */
+	if (IS_ENCRYPTED(inode_in) != IS_ENCRYPTED(inode_out))
+		return -EINVAL;
+
 	/* Don't make the dst file partly checksummed */
 	if ((BTRFS_I(inode_in)->flags & BTRFS_INODE_NODATASUM) !=
 	    (BTRFS_I(inode_out)->flags & BTRFS_INODE_NODATASUM)) {