2017-06-15 02:37:39 +08:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved.
|
|
|
|
* Copyright (c) 2016-2017, Dave Watson <davejwatson@fb.com>. All rights reserved.
|
|
|
|
*
|
|
|
|
* This software is available to you under a choice of one of two
|
|
|
|
* licenses. You may choose to be licensed under the terms of the GNU
|
|
|
|
* General Public License (GPL) Version 2, available from the file
|
|
|
|
* COPYING in the main directory of this source tree, or the
|
|
|
|
* OpenIB.org BSD license below:
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or
|
|
|
|
* without modification, are permitted provided that the following
|
|
|
|
* conditions are met:
|
|
|
|
*
|
|
|
|
* - Redistributions of source code must retain the above
|
|
|
|
* copyright notice, this list of conditions and the following
|
|
|
|
* disclaimer.
|
|
|
|
*
|
|
|
|
* - Redistributions in binary form must reproduce the above
|
|
|
|
* copyright notice, this list of conditions and the following
|
|
|
|
* disclaimer in the documentation and/or other materials
|
|
|
|
* provided with the distribution.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
|
|
|
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
|
|
|
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
|
|
|
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
|
|
|
|
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
|
|
|
|
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
|
|
|
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
|
|
* SOFTWARE.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef _TLS_OFFLOAD_H
|
|
|
|
#define _TLS_OFFLOAD_H
|
|
|
|
|
|
|
|
#include <linux/types.h>
|
2017-11-14 11:30:11 +08:00
|
|
|
#include <asm/byteorder.h>
|
2018-02-01 00:04:37 +08:00
|
|
|
#include <linux/crypto.h>
|
2017-11-14 11:30:11 +08:00
|
|
|
#include <linux/socket.h>
|
|
|
|
#include <linux/tcp.h>
|
2018-10-13 08:45:59 +08:00
|
|
|
#include <linux/skmsg.h>
|
|
|
|
|
2017-11-14 11:30:11 +08:00
|
|
|
#include <net/tcp.h>
|
tls: RX path for ktls
Add rx path for tls software implementation.
recvmsg, splice_read, and poll implemented.
An additional sockopt TLS_RX is added, with the same interface as
TLS_TX. Either TLX_RX or TLX_TX may be provided separately, or
together (with two different setsockopt calls with appropriate keys).
Control messages are passed via CMSG in a similar way to transmit.
If no cmsg buffer is passed, then only application data records
will be passed to userspace, and EIO is returned for other types of
alerts.
EBADMSG is passed for decryption errors, and EMSGSIZE is passed for
framing too big, and EBADMSG for framing too small (matching openssl
semantics). EINVAL is returned for TLS versions that do not match the
original setsockopt call. All are unrecoverable.
strparser is used to parse TLS framing. Decryption is done directly
in to userspace buffers if they are large enough to support it, otherwise
sk_cow_data is called (similar to ipsec), and buffers are decrypted in
place and copied. splice_read always decrypts in place, since no
buffers are provided to decrypt in to.
sk_poll is overridden, and only returns POLLIN if a full TLS message is
received. Otherwise we wait for strparser to finish reading a full frame.
Actual decryption is only done during recvmsg or splice_read calls.
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 01:10:35 +08:00
|
|
|
#include <net/strparser.h>
|
2018-09-21 12:16:13 +08:00
|
|
|
#include <crypto/aead.h>
|
2017-06-15 02:37:39 +08:00
|
|
|
#include <uapi/linux/tls.h>
|
|
|
|
|
|
|
|
|
|
|
|
/* Maximum data size carried in a TLS record */
|
|
|
|
#define TLS_MAX_PAYLOAD_SIZE ((size_t)1 << 14)
|
|
|
|
|
|
|
|
#define TLS_HEADER_SIZE 5
|
|
|
|
#define TLS_NONCE_OFFSET TLS_HEADER_SIZE
|
|
|
|
|
|
|
|
#define TLS_CRYPTO_INFO_READY(info) ((info)->cipher_type)
|
|
|
|
|
|
|
|
#define TLS_RECORD_TYPE_DATA 0x17
|
|
|
|
|
|
|
|
#define TLS_AAD_SPACE_SIZE 13
|
2018-04-01 00:11:52 +08:00
|
|
|
#define TLS_DEVICE_NAME_MAX 32
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This structure defines the routines for Inline TLS driver.
|
|
|
|
* The following routines are optional and filled with a
|
|
|
|
* null pointer if not defined.
|
|
|
|
*
|
|
|
|
* @name: Its the name of registered Inline tls device
|
|
|
|
* @dev_list: Inline tls device list
|
|
|
|
* int (*feature)(struct tls_device *device);
|
|
|
|
* Called to return Inline TLS driver capability
|
|
|
|
*
|
|
|
|
* int (*hash)(struct tls_device *device, struct sock *sk);
|
|
|
|
* This function sets Inline driver for listen and program
|
|
|
|
* device specific functioanlity as required
|
|
|
|
*
|
|
|
|
* void (*unhash)(struct tls_device *device, struct sock *sk);
|
|
|
|
* This function cleans listen state set by Inline TLS driver
|
2018-12-11 18:20:09 +08:00
|
|
|
*
|
|
|
|
* void (*release)(struct kref *kref);
|
|
|
|
* Release the registered device and allocated resources
|
|
|
|
* @kref: Number of reference to tls_device
|
2018-04-01 00:11:52 +08:00
|
|
|
*/
|
|
|
|
struct tls_device {
|
|
|
|
char name[TLS_DEVICE_NAME_MAX];
|
|
|
|
struct list_head dev_list;
|
|
|
|
int (*feature)(struct tls_device *device);
|
|
|
|
int (*hash)(struct tls_device *device, struct sock *sk);
|
|
|
|
void (*unhash)(struct tls_device *device, struct sock *sk);
|
2018-12-11 18:20:09 +08:00
|
|
|
void (*release)(struct kref *kref);
|
|
|
|
struct kref kref;
|
2018-04-01 00:11:52 +08:00
|
|
|
};
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-07-13 19:33:43 +08:00
|
|
|
enum {
|
|
|
|
TLS_BASE,
|
|
|
|
TLS_SW,
|
|
|
|
#ifdef CONFIG_TLS_DEVICE
|
|
|
|
TLS_HW,
|
|
|
|
#endif
|
|
|
|
TLS_HW_RECORD,
|
|
|
|
TLS_NUM_CONFIG,
|
|
|
|
};
|
|
|
|
|
2018-09-21 12:16:13 +08:00
|
|
|
/* TLS records are maintained in 'struct tls_rec'. It stores the memory pages
|
|
|
|
* allocated or mapped for each TLS record. After encryption, the records are
|
|
|
|
* stores in a linked list.
|
|
|
|
*/
|
|
|
|
struct tls_rec {
|
|
|
|
struct list_head list;
|
net/tls: Fixed race condition in async encryption
On processors with multi-engine crypto accelerators, it is possible that
multiple records get encrypted in parallel and their encryption
completion is notified to different cpus in multicore processor. This
leads to the situation where tls_encrypt_done() starts executing in
parallel on different cores. In current implementation, encrypted
records are queued to tx_ready_list in tls_encrypt_done(). This requires
addition to linked list 'tx_ready_list' to be protected. As
tls_decrypt_done() could be executing in irq content, it is not possible
to protect linked list addition operation using a lock.
To fix the problem, we remove linked list addition operation from the
irq context. We do tx_ready_list addition/removal operation from
application context only and get rid of possible multiple access to
the linked list. Before starting encryption on the record, we add it to
the tail of tx_ready_list. To prevent tls_tx_records() from transmitting
it, we mark the record with a new flag 'tx_ready' in 'struct tls_rec'.
When record encryption gets completed, tls_encrypt_done() has to only
update the 'tx_ready' flag to true & linked list add operation is not
required.
The changed logic brings some other side benefits. Since the records
are always submitted in tls sequence number order for encryption, the
tx_ready_list always remains sorted and addition of new records to it
does not have to traverse the linked list.
Lastly, we renamed tx_ready_list in 'struct tls_sw_context_tx' to
'tx_list'. This is because now, the some of the records at the tail are
not ready to transmit.
Fixes: a42055e8d2c3 ("net/tls: Add support for async encryption")
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-24 18:05:56 +08:00
|
|
|
int tx_ready;
|
2018-09-21 12:16:13 +08:00
|
|
|
int tx_flags;
|
2018-09-30 10:34:35 +08:00
|
|
|
int inplace_crypto;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-10-13 08:45:59 +08:00
|
|
|
struct sk_msg msg_plaintext;
|
|
|
|
struct sk_msg msg_encrypted;
|
2018-09-21 12:16:13 +08:00
|
|
|
|
2018-10-13 08:45:59 +08:00
|
|
|
/* AAD | msg_plaintext.sg.data | sg_tag */
|
|
|
|
struct scatterlist sg_aead_in[2];
|
|
|
|
/* AAD | msg_encrypted.sg.data (data contains overhead for hdr & iv & tag) */
|
|
|
|
struct scatterlist sg_aead_out[2];
|
2018-09-21 12:16:13 +08:00
|
|
|
|
|
|
|
char aad_space[TLS_AAD_SPACE_SIZE];
|
2019-01-27 08:57:38 +08:00
|
|
|
u8 iv_data[TLS_CIPHER_AES_GCM_128_IV_SIZE +
|
|
|
|
TLS_CIPHER_AES_GCM_128_SALT_SIZE];
|
2018-09-21 12:16:13 +08:00
|
|
|
struct aead_request aead_req;
|
|
|
|
u8 aead_req_ctx[];
|
|
|
|
};
|
|
|
|
|
|
|
|
struct tx_work {
|
|
|
|
struct delayed_work work;
|
|
|
|
struct sock *sk;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct tls_sw_context_tx {
|
|
|
|
struct crypto_aead *aead_send;
|
|
|
|
struct crypto_wait async_wait;
|
|
|
|
struct tx_work tx_work;
|
|
|
|
struct tls_rec *open_rec;
|
net/tls: Fixed race condition in async encryption
On processors with multi-engine crypto accelerators, it is possible that
multiple records get encrypted in parallel and their encryption
completion is notified to different cpus in multicore processor. This
leads to the situation where tls_encrypt_done() starts executing in
parallel on different cores. In current implementation, encrypted
records are queued to tx_ready_list in tls_encrypt_done(). This requires
addition to linked list 'tx_ready_list' to be protected. As
tls_decrypt_done() could be executing in irq content, it is not possible
to protect linked list addition operation using a lock.
To fix the problem, we remove linked list addition operation from the
irq context. We do tx_ready_list addition/removal operation from
application context only and get rid of possible multiple access to
the linked list. Before starting encryption on the record, we add it to
the tail of tx_ready_list. To prevent tls_tx_records() from transmitting
it, we mark the record with a new flag 'tx_ready' in 'struct tls_rec'.
When record encryption gets completed, tls_encrypt_done() has to only
update the 'tx_ready' flag to true & linked list add operation is not
required.
The changed logic brings some other side benefits. Since the records
are always submitted in tls sequence number order for encryption, the
tx_ready_list always remains sorted and addition of new records to it
does not have to traverse the linked list.
Lastly, we renamed tx_ready_list in 'struct tls_sw_context_tx' to
'tx_list'. This is because now, the some of the records at the tail are
not ready to transmit.
Fixes: a42055e8d2c3 ("net/tls: Add support for async encryption")
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-24 18:05:56 +08:00
|
|
|
struct list_head tx_list;
|
2018-09-21 12:16:13 +08:00
|
|
|
atomic_t encrypt_pending;
|
|
|
|
int async_notify;
|
|
|
|
|
|
|
|
#define BIT_TX_SCHEDULED 0
|
|
|
|
unsigned long tx_bitmask;
|
2017-06-15 02:37:39 +08:00
|
|
|
};
|
|
|
|
|
2018-04-30 15:16:15 +08:00
|
|
|
struct tls_sw_context_rx {
|
|
|
|
struct crypto_aead *aead_recv;
|
|
|
|
struct crypto_wait async_wait;
|
|
|
|
struct strparser strp;
|
2019-01-16 18:40:16 +08:00
|
|
|
struct sk_buff_head rx_list; /* list of decrypted 'data' records */
|
2018-04-30 15:16:15 +08:00
|
|
|
void (*saved_data_ready)(struct sock *sk);
|
2018-10-13 08:46:00 +08:00
|
|
|
|
2018-04-30 15:16:15 +08:00
|
|
|
struct sk_buff *recv_pkt;
|
|
|
|
u8 control;
|
2019-01-16 18:40:16 +08:00
|
|
|
int async_capable;
|
2018-04-30 15:16:15 +08:00
|
|
|
bool decrypted;
|
2018-08-29 17:56:55 +08:00
|
|
|
atomic_t decrypt_pending;
|
|
|
|
bool async_notify;
|
|
|
|
};
|
|
|
|
|
2018-04-30 15:16:16 +08:00
|
|
|
struct tls_record_info {
|
|
|
|
struct list_head list;
|
|
|
|
u32 end_seq;
|
|
|
|
int len;
|
|
|
|
int num_frags;
|
|
|
|
skb_frag_t frags[MAX_SKB_FRAGS];
|
|
|
|
};
|
|
|
|
|
2018-07-13 19:33:39 +08:00
|
|
|
struct tls_offload_context_tx {
|
2018-04-30 15:16:16 +08:00
|
|
|
struct crypto_aead *aead_send;
|
|
|
|
spinlock_t lock; /* protects records list */
|
|
|
|
struct list_head records_list;
|
|
|
|
struct tls_record_info *open_record;
|
|
|
|
struct tls_record_info *retransmit_hint;
|
|
|
|
u64 hint_record_sn;
|
|
|
|
u64 unacked_record_sn;
|
|
|
|
|
|
|
|
struct scatterlist sg_tx_data[MAX_SKB_FRAGS];
|
|
|
|
void (*sk_destruct)(struct sock *sk);
|
|
|
|
u8 driver_state[];
|
|
|
|
/* The TLS layer reserves room for driver specific state
|
|
|
|
* Currently the belief is that there is not enough
|
|
|
|
* driver specific state to justify another layer of indirection
|
|
|
|
*/
|
|
|
|
#define TLS_DRIVER_STATE_SIZE (max_t(size_t, 8, sizeof(void *)))
|
|
|
|
};
|
|
|
|
|
2018-07-13 19:33:39 +08:00
|
|
|
#define TLS_OFFLOAD_CONTEXT_SIZE_TX \
|
|
|
|
(ALIGN(sizeof(struct tls_offload_context_tx), sizeof(void *)) + \
|
2018-04-30 15:16:16 +08:00
|
|
|
TLS_DRIVER_STATE_SIZE)
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
enum {
|
|
|
|
TLS_PENDING_CLOSED_RECORD
|
|
|
|
};
|
|
|
|
|
2018-03-23 01:10:06 +08:00
|
|
|
struct cipher_context {
|
|
|
|
u16 prepend_size;
|
|
|
|
u16 tag_size;
|
|
|
|
u16 overhead_size;
|
|
|
|
u16 iv_size;
|
|
|
|
char *iv;
|
|
|
|
u16 rec_seq_size;
|
|
|
|
char *rec_seq;
|
|
|
|
};
|
|
|
|
|
2018-09-12 23:44:42 +08:00
|
|
|
union tls_crypto_context {
|
|
|
|
struct tls_crypto_info info;
|
|
|
|
struct tls12_crypto_info_aes_gcm_128 aes_gcm_128;
|
|
|
|
};
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
struct tls_context {
|
2018-09-12 23:44:42 +08:00
|
|
|
union tls_crypto_context crypto_send;
|
|
|
|
union tls_crypto_context crypto_recv;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-04-30 15:16:15 +08:00
|
|
|
struct list_head list;
|
|
|
|
struct net_device *netdev;
|
|
|
|
refcount_t refcount;
|
|
|
|
|
|
|
|
void *priv_ctx_tx;
|
|
|
|
void *priv_ctx_rx;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-04-30 15:16:15 +08:00
|
|
|
u8 tx_conf:3;
|
|
|
|
u8 rx_conf:3;
|
2017-11-13 16:22:45 +08:00
|
|
|
|
2018-03-23 01:10:06 +08:00
|
|
|
struct cipher_context tx;
|
tls: RX path for ktls
Add rx path for tls software implementation.
recvmsg, splice_read, and poll implemented.
An additional sockopt TLS_RX is added, with the same interface as
TLS_TX. Either TLX_RX or TLX_TX may be provided separately, or
together (with two different setsockopt calls with appropriate keys).
Control messages are passed via CMSG in a similar way to transmit.
If no cmsg buffer is passed, then only application data records
will be passed to userspace, and EIO is returned for other types of
alerts.
EBADMSG is passed for decryption errors, and EMSGSIZE is passed for
framing too big, and EBADMSG for framing too small (matching openssl
semantics). EINVAL is returned for TLS versions that do not match the
original setsockopt call. All are unrecoverable.
strparser is used to parse TLS framing. Decryption is done directly
in to userspace buffers if they are large enough to support it, otherwise
sk_cow_data is called (similar to ipsec), and buffers are decrypted in
place and copied. splice_read always decrypts in place, since no
buffers are provided to decrypt in to.
sk_poll is overridden, and only returns POLLIN if a full TLS message is
received. Otherwise we wait for strparser to finish reading a full frame.
Actual decryption is only done during recvmsg or splice_read calls.
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 01:10:35 +08:00
|
|
|
struct cipher_context rx;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
|
|
|
struct scatterlist *partially_sent_record;
|
|
|
|
u16 partially_sent_offset;
|
2018-09-21 12:16:13 +08:00
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
unsigned long flags;
|
2018-05-02 04:05:39 +08:00
|
|
|
bool in_tcp_sendpages;
|
2018-10-13 08:45:59 +08:00
|
|
|
bool pending_open_record_frags;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
|
|
|
int (*push_pending_record)(struct sock *sk, int flags);
|
|
|
|
|
|
|
|
void (*sk_write_space)(struct sock *sk);
|
2018-07-13 19:33:43 +08:00
|
|
|
void (*sk_destruct)(struct sock *sk);
|
2017-06-15 02:37:39 +08:00
|
|
|
void (*sk_proto_close)(struct sock *sk, long timeout);
|
|
|
|
|
|
|
|
int (*setsockopt)(struct sock *sk, int level,
|
|
|
|
int optname, char __user *optval,
|
|
|
|
unsigned int optlen);
|
|
|
|
int (*getsockopt)(struct sock *sk, int level,
|
|
|
|
int optname, char __user *optval,
|
|
|
|
int __user *optlen);
|
2018-04-01 00:11:52 +08:00
|
|
|
int (*hash)(struct sock *sk);
|
|
|
|
void (*unhash)(struct sock *sk);
|
2017-06-15 02:37:39 +08:00
|
|
|
};
|
|
|
|
|
2018-07-13 19:33:43 +08:00
|
|
|
struct tls_offload_context_rx {
|
|
|
|
/* sw must be the first member of tls_offload_context_rx */
|
|
|
|
struct tls_sw_context_rx sw;
|
|
|
|
atomic64_t resync_req;
|
|
|
|
u8 driver_state[];
|
|
|
|
/* The TLS layer reserves room for driver specific state
|
|
|
|
* Currently the belief is that there is not enough
|
|
|
|
* driver specific state to justify another layer of indirection
|
|
|
|
*/
|
|
|
|
};
|
|
|
|
|
|
|
|
#define TLS_OFFLOAD_CONTEXT_SIZE_RX \
|
|
|
|
(ALIGN(sizeof(struct tls_offload_context_rx), sizeof(void *)) + \
|
|
|
|
TLS_DRIVER_STATE_SIZE)
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
int wait_on_pending_writer(struct sock *sk, long *timeo);
|
|
|
|
int tls_sk_query(struct sock *sk, int optname, char __user *optval,
|
|
|
|
int __user *optlen);
|
|
|
|
int tls_sk_attach(struct sock *sk, int optname, char __user *optval,
|
|
|
|
unsigned int optlen);
|
|
|
|
|
tls: RX path for ktls
Add rx path for tls software implementation.
recvmsg, splice_read, and poll implemented.
An additional sockopt TLS_RX is added, with the same interface as
TLS_TX. Either TLX_RX or TLX_TX may be provided separately, or
together (with two different setsockopt calls with appropriate keys).
Control messages are passed via CMSG in a similar way to transmit.
If no cmsg buffer is passed, then only application data records
will be passed to userspace, and EIO is returned for other types of
alerts.
EBADMSG is passed for decryption errors, and EMSGSIZE is passed for
framing too big, and EBADMSG for framing too small (matching openssl
semantics). EINVAL is returned for TLS versions that do not match the
original setsockopt call. All are unrecoverable.
strparser is used to parse TLS framing. Decryption is done directly
in to userspace buffers if they are large enough to support it, otherwise
sk_cow_data is called (similar to ipsec), and buffers are decrypted in
place and copied. splice_read always decrypts in place, since no
buffers are provided to decrypt in to.
sk_poll is overridden, and only returns POLLIN if a full TLS message is
received. Otherwise we wait for strparser to finish reading a full frame.
Actual decryption is only done during recvmsg or splice_read calls.
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 01:10:35 +08:00
|
|
|
int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx, int tx);
|
2017-06-15 02:37:39 +08:00
|
|
|
int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size);
|
|
|
|
int tls_sw_sendpage(struct sock *sk, struct page *page,
|
|
|
|
int offset, size_t size, int flags);
|
|
|
|
void tls_sw_close(struct sock *sk, long timeout);
|
2018-04-30 15:16:15 +08:00
|
|
|
void tls_sw_free_resources_tx(struct sock *sk);
|
|
|
|
void tls_sw_free_resources_rx(struct sock *sk);
|
2018-07-13 19:33:41 +08:00
|
|
|
void tls_sw_release_resources_rx(struct sock *sk);
|
tls: RX path for ktls
Add rx path for tls software implementation.
recvmsg, splice_read, and poll implemented.
An additional sockopt TLS_RX is added, with the same interface as
TLS_TX. Either TLX_RX or TLX_TX may be provided separately, or
together (with two different setsockopt calls with appropriate keys).
Control messages are passed via CMSG in a similar way to transmit.
If no cmsg buffer is passed, then only application data records
will be passed to userspace, and EIO is returned for other types of
alerts.
EBADMSG is passed for decryption errors, and EMSGSIZE is passed for
framing too big, and EBADMSG for framing too small (matching openssl
semantics). EINVAL is returned for TLS versions that do not match the
original setsockopt call. All are unrecoverable.
strparser is used to parse TLS framing. Decryption is done directly
in to userspace buffers if they are large enough to support it, otherwise
sk_cow_data is called (similar to ipsec), and buffers are decrypted in
place and copied. splice_read always decrypts in place, since no
buffers are provided to decrypt in to.
sk_poll is overridden, and only returns POLLIN if a full TLS message is
received. Otherwise we wait for strparser to finish reading a full frame.
Actual decryption is only done during recvmsg or splice_read calls.
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 01:10:35 +08:00
|
|
|
int tls_sw_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
|
|
|
|
int nonblock, int flags, int *addr_len);
|
2018-10-13 08:46:00 +08:00
|
|
|
bool tls_sw_stream_read(const struct sock *sk);
|
tls: RX path for ktls
Add rx path for tls software implementation.
recvmsg, splice_read, and poll implemented.
An additional sockopt TLS_RX is added, with the same interface as
TLS_TX. Either TLX_RX or TLX_TX may be provided separately, or
together (with two different setsockopt calls with appropriate keys).
Control messages are passed via CMSG in a similar way to transmit.
If no cmsg buffer is passed, then only application data records
will be passed to userspace, and EIO is returned for other types of
alerts.
EBADMSG is passed for decryption errors, and EMSGSIZE is passed for
framing too big, and EBADMSG for framing too small (matching openssl
semantics). EINVAL is returned for TLS versions that do not match the
original setsockopt call. All are unrecoverable.
strparser is used to parse TLS framing. Decryption is done directly
in to userspace buffers if they are large enough to support it, otherwise
sk_cow_data is called (similar to ipsec), and buffers are decrypted in
place and copied. splice_read always decrypts in place, since no
buffers are provided to decrypt in to.
sk_poll is overridden, and only returns POLLIN if a full TLS message is
received. Otherwise we wait for strparser to finish reading a full frame.
Actual decryption is only done during recvmsg or splice_read calls.
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 01:10:35 +08:00
|
|
|
ssize_t tls_sw_splice_read(struct socket *sock, loff_t *ppos,
|
|
|
|
struct pipe_inode_info *pipe,
|
|
|
|
size_t len, unsigned int flags);
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-04-30 15:16:16 +08:00
|
|
|
int tls_set_device_offload(struct sock *sk, struct tls_context *ctx);
|
|
|
|
int tls_device_sendmsg(struct sock *sk, struct msghdr *msg, size_t size);
|
|
|
|
int tls_device_sendpage(struct sock *sk, struct page *page,
|
|
|
|
int offset, size_t size, int flags);
|
|
|
|
void tls_device_sk_destruct(struct sock *sk);
|
|
|
|
void tls_device_init(void);
|
|
|
|
void tls_device_cleanup(void);
|
2018-09-21 12:16:13 +08:00
|
|
|
int tls_tx_records(struct sock *sk, int flags);
|
2018-04-30 15:16:16 +08:00
|
|
|
|
2018-07-13 19:33:39 +08:00
|
|
|
struct tls_record_info *tls_get_record(struct tls_offload_context_tx *context,
|
2018-04-30 15:16:16 +08:00
|
|
|
u32 seq, u64 *p_record_sn);
|
|
|
|
|
|
|
|
static inline bool tls_record_is_start_marker(struct tls_record_info *rec)
|
|
|
|
{
|
|
|
|
return rec->len == 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline u32 tls_record_start_seq(struct tls_record_info *rec)
|
|
|
|
{
|
|
|
|
return rec->end_seq - rec->len;
|
|
|
|
}
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-04-30 15:16:16 +08:00
|
|
|
void tls_sk_destruct(struct sock *sk, struct tls_context *ctx);
|
2017-06-15 02:37:39 +08:00
|
|
|
int tls_push_sg(struct sock *sk, struct tls_context *ctx,
|
|
|
|
struct scatterlist *sg, u16 first_offset,
|
|
|
|
int flags);
|
2018-09-21 12:16:13 +08:00
|
|
|
int tls_push_partial_record(struct sock *sk, struct tls_context *ctx,
|
|
|
|
int flags);
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
int tls_push_pending_closed_record(struct sock *sk, struct tls_context *ctx,
|
|
|
|
int flags, long *timeo);
|
|
|
|
|
|
|
|
static inline bool tls_is_pending_closed_record(struct tls_context *ctx)
|
|
|
|
{
|
|
|
|
return test_bit(TLS_PENDING_CLOSED_RECORD, &ctx->flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int tls_complete_pending_work(struct sock *sk,
|
|
|
|
struct tls_context *ctx,
|
|
|
|
int flags, long *timeo)
|
|
|
|
{
|
|
|
|
int rc = 0;
|
|
|
|
|
|
|
|
if (unlikely(sk->sk_write_pending))
|
|
|
|
rc = wait_on_pending_writer(sk, timeo);
|
|
|
|
|
|
|
|
if (!rc && tls_is_pending_closed_record(ctx))
|
|
|
|
rc = tls_push_pending_closed_record(sk, ctx, flags, timeo);
|
|
|
|
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool tls_is_partially_sent_record(struct tls_context *ctx)
|
|
|
|
{
|
|
|
|
return !!ctx->partially_sent_record;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool tls_is_pending_open_record(struct tls_context *tls_ctx)
|
|
|
|
{
|
|
|
|
return tls_ctx->pending_open_record_frags;
|
|
|
|
}
|
|
|
|
|
net/tls: Fixed race condition in async encryption
On processors with multi-engine crypto accelerators, it is possible that
multiple records get encrypted in parallel and their encryption
completion is notified to different cpus in multicore processor. This
leads to the situation where tls_encrypt_done() starts executing in
parallel on different cores. In current implementation, encrypted
records are queued to tx_ready_list in tls_encrypt_done(). This requires
addition to linked list 'tx_ready_list' to be protected. As
tls_decrypt_done() could be executing in irq content, it is not possible
to protect linked list addition operation using a lock.
To fix the problem, we remove linked list addition operation from the
irq context. We do tx_ready_list addition/removal operation from
application context only and get rid of possible multiple access to
the linked list. Before starting encryption on the record, we add it to
the tail of tx_ready_list. To prevent tls_tx_records() from transmitting
it, we mark the record with a new flag 'tx_ready' in 'struct tls_rec'.
When record encryption gets completed, tls_encrypt_done() has to only
update the 'tx_ready' flag to true & linked list add operation is not
required.
The changed logic brings some other side benefits. Since the records
are always submitted in tls sequence number order for encryption, the
tx_ready_list always remains sorted and addition of new records to it
does not have to traverse the linked list.
Lastly, we renamed tx_ready_list in 'struct tls_sw_context_tx' to
'tx_list'. This is because now, the some of the records at the tail are
not ready to transmit.
Fixes: a42055e8d2c3 ("net/tls: Add support for async encryption")
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-24 18:05:56 +08:00
|
|
|
static inline bool is_tx_ready(struct tls_sw_context_tx *ctx)
|
2018-09-21 12:16:13 +08:00
|
|
|
{
|
|
|
|
struct tls_rec *rec;
|
|
|
|
|
net/tls: Fixed race condition in async encryption
On processors with multi-engine crypto accelerators, it is possible that
multiple records get encrypted in parallel and their encryption
completion is notified to different cpus in multicore processor. This
leads to the situation where tls_encrypt_done() starts executing in
parallel on different cores. In current implementation, encrypted
records are queued to tx_ready_list in tls_encrypt_done(). This requires
addition to linked list 'tx_ready_list' to be protected. As
tls_decrypt_done() could be executing in irq content, it is not possible
to protect linked list addition operation using a lock.
To fix the problem, we remove linked list addition operation from the
irq context. We do tx_ready_list addition/removal operation from
application context only and get rid of possible multiple access to
the linked list. Before starting encryption on the record, we add it to
the tail of tx_ready_list. To prevent tls_tx_records() from transmitting
it, we mark the record with a new flag 'tx_ready' in 'struct tls_rec'.
When record encryption gets completed, tls_encrypt_done() has to only
update the 'tx_ready' flag to true & linked list add operation is not
required.
The changed logic brings some other side benefits. Since the records
are always submitted in tls sequence number order for encryption, the
tx_ready_list always remains sorted and addition of new records to it
does not have to traverse the linked list.
Lastly, we renamed tx_ready_list in 'struct tls_sw_context_tx' to
'tx_list'. This is because now, the some of the records at the tail are
not ready to transmit.
Fixes: a42055e8d2c3 ("net/tls: Add support for async encryption")
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-24 18:05:56 +08:00
|
|
|
rec = list_first_entry(&ctx->tx_list, struct tls_rec, list);
|
2018-09-21 12:16:13 +08:00
|
|
|
if (!rec)
|
|
|
|
return false;
|
|
|
|
|
net/tls: Fixed race condition in async encryption
On processors with multi-engine crypto accelerators, it is possible that
multiple records get encrypted in parallel and their encryption
completion is notified to different cpus in multicore processor. This
leads to the situation where tls_encrypt_done() starts executing in
parallel on different cores. In current implementation, encrypted
records are queued to tx_ready_list in tls_encrypt_done(). This requires
addition to linked list 'tx_ready_list' to be protected. As
tls_decrypt_done() could be executing in irq content, it is not possible
to protect linked list addition operation using a lock.
To fix the problem, we remove linked list addition operation from the
irq context. We do tx_ready_list addition/removal operation from
application context only and get rid of possible multiple access to
the linked list. Before starting encryption on the record, we add it to
the tail of tx_ready_list. To prevent tls_tx_records() from transmitting
it, we mark the record with a new flag 'tx_ready' in 'struct tls_rec'.
When record encryption gets completed, tls_encrypt_done() has to only
update the 'tx_ready' flag to true & linked list add operation is not
required.
The changed logic brings some other side benefits. Since the records
are always submitted in tls sequence number order for encryption, the
tx_ready_list always remains sorted and addition of new records to it
does not have to traverse the linked list.
Lastly, we renamed tx_ready_list in 'struct tls_sw_context_tx' to
'tx_list'. This is because now, the some of the records at the tail are
not ready to transmit.
Fixes: a42055e8d2c3 ("net/tls: Add support for async encryption")
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-24 18:05:56 +08:00
|
|
|
return READ_ONCE(rec->tx_ready);
|
2018-09-21 12:16:13 +08:00
|
|
|
}
|
|
|
|
|
2018-07-13 19:33:43 +08:00
|
|
|
struct sk_buff *
|
|
|
|
tls_validate_xmit_skb(struct sock *sk, struct net_device *dev,
|
|
|
|
struct sk_buff *skb);
|
|
|
|
|
2018-04-30 15:16:16 +08:00
|
|
|
static inline bool tls_is_sk_tx_device_offloaded(struct sock *sk)
|
|
|
|
{
|
2018-07-13 19:33:43 +08:00
|
|
|
#ifdef CONFIG_SOCK_VALIDATE_XMIT
|
|
|
|
return sk_fullsock(sk) &
|
|
|
|
(smp_load_acquire(&sk->sk_validate_xmit_skb) ==
|
|
|
|
&tls_validate_xmit_skb);
|
|
|
|
#else
|
|
|
|
return false;
|
|
|
|
#endif
|
2018-04-30 15:16:16 +08:00
|
|
|
}
|
|
|
|
|
2018-03-23 01:10:15 +08:00
|
|
|
static inline void tls_err_abort(struct sock *sk, int err)
|
2017-06-15 02:37:39 +08:00
|
|
|
{
|
2018-03-23 01:10:15 +08:00
|
|
|
sk->sk_err = err;
|
2017-06-15 02:37:39 +08:00
|
|
|
sk->sk_error_report(sk);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool tls_bigint_increment(unsigned char *seq, int len)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = len - 1; i >= 0; i--) {
|
|
|
|
++seq[i];
|
|
|
|
if (seq[i] != 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return (i == -1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void tls_advance_record_sn(struct sock *sk,
|
2018-03-23 01:10:06 +08:00
|
|
|
struct cipher_context *ctx)
|
2017-06-15 02:37:39 +08:00
|
|
|
{
|
|
|
|
if (tls_bigint_increment(ctx->rec_seq, ctx->rec_seq_size))
|
2018-03-23 01:10:15 +08:00
|
|
|
tls_err_abort(sk, EBADMSG);
|
2017-06-15 02:37:39 +08:00
|
|
|
tls_bigint_increment(ctx->iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE,
|
|
|
|
ctx->iv_size);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void tls_fill_prepend(struct tls_context *ctx,
|
|
|
|
char *buf,
|
|
|
|
size_t plaintext_len,
|
|
|
|
unsigned char record_type)
|
|
|
|
{
|
2018-03-23 01:10:06 +08:00
|
|
|
size_t pkt_len, iv_size = ctx->tx.iv_size;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-03-23 01:10:06 +08:00
|
|
|
pkt_len = plaintext_len + iv_size + ctx->tx.tag_size;
|
2017-06-15 02:37:39 +08:00
|
|
|
|
|
|
|
/* we cover nonce explicit here as well, so buf should be of
|
|
|
|
* size KTLS_DTLS_HEADER_SIZE + KTLS_DTLS_NONCE_EXPLICIT_SIZE
|
|
|
|
*/
|
|
|
|
buf[0] = record_type;
|
2018-09-12 23:44:42 +08:00
|
|
|
buf[1] = TLS_VERSION_MINOR(ctx->crypto_send.info.version);
|
|
|
|
buf[2] = TLS_VERSION_MAJOR(ctx->crypto_send.info.version);
|
2017-06-15 02:37:39 +08:00
|
|
|
/* we can use IV for nonce explicit according to spec */
|
|
|
|
buf[3] = pkt_len >> 8;
|
|
|
|
buf[4] = pkt_len & 0xFF;
|
|
|
|
memcpy(buf + TLS_NONCE_OFFSET,
|
2018-03-23 01:10:06 +08:00
|
|
|
ctx->tx.iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, iv_size);
|
2017-06-15 02:37:39 +08:00
|
|
|
}
|
|
|
|
|
2017-11-13 16:22:47 +08:00
|
|
|
static inline void tls_make_aad(char *buf,
|
|
|
|
size_t size,
|
|
|
|
char *record_sequence,
|
|
|
|
int record_sequence_size,
|
|
|
|
unsigned char record_type)
|
|
|
|
{
|
|
|
|
memcpy(buf, record_sequence, record_sequence_size);
|
|
|
|
|
|
|
|
buf[8] = record_type;
|
|
|
|
buf[9] = TLS_1_2_VERSION_MAJOR;
|
|
|
|
buf[10] = TLS_1_2_VERSION_MINOR;
|
|
|
|
buf[11] = size >> 8;
|
|
|
|
buf[12] = size & 0xFF;
|
|
|
|
}
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
static inline struct tls_context *tls_get_ctx(const struct sock *sk)
|
|
|
|
{
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
|
|
|
return icsk->icsk_ulp_data;
|
|
|
|
}
|
|
|
|
|
2018-04-30 15:16:15 +08:00
|
|
|
static inline struct tls_sw_context_rx *tls_sw_ctx_rx(
|
|
|
|
const struct tls_context *tls_ctx)
|
|
|
|
{
|
|
|
|
return (struct tls_sw_context_rx *)tls_ctx->priv_ctx_rx;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct tls_sw_context_tx *tls_sw_ctx_tx(
|
2017-06-15 02:37:39 +08:00
|
|
|
const struct tls_context *tls_ctx)
|
|
|
|
{
|
2018-04-30 15:16:15 +08:00
|
|
|
return (struct tls_sw_context_tx *)tls_ctx->priv_ctx_tx;
|
2017-06-15 02:37:39 +08:00
|
|
|
}
|
|
|
|
|
2018-07-13 19:33:39 +08:00
|
|
|
static inline struct tls_offload_context_tx *
|
|
|
|
tls_offload_ctx_tx(const struct tls_context *tls_ctx)
|
2017-06-15 02:37:39 +08:00
|
|
|
{
|
2018-07-13 19:33:39 +08:00
|
|
|
return (struct tls_offload_context_tx *)tls_ctx->priv_ctx_tx;
|
2017-06-15 02:37:39 +08:00
|
|
|
}
|
|
|
|
|
bpf: sk_msg, sock{map|hash} redirect through ULP
A sockmap program that redirects through a kTLS ULP enabled socket
will not work correctly because the ULP layer is skipped. This
fixes the behavior to call through the ULP layer on redirect to
ensure any operations required on the data stream at the ULP layer
continue to be applied.
To do this we add an internal flag MSG_SENDPAGE_NOPOLICY to avoid
calling the BPF layer on a redirected message. This is
required to avoid calling the BPF layer multiple times (possibly
recursively) which is not the current/expected behavior without
ULPs. In the future we may add a redirect flag if users _do_
want the policy applied again but this would need to work for both
ULP and non-ULP sockets and be opt-in to avoid breaking existing
programs.
Also to avoid polluting the flag space with an internal flag we
reuse the flag space overlapping MSG_SENDPAGE_NOPOLICY with
MSG_WAITFORONE. Here WAITFORONE is specific to recv path and
SENDPAGE_NOPOLICY is only used for sendpage hooks. The last thing
to verify is user space API is masked correctly to ensure the flag
can not be set by user. (Note this needs to be true regardless
because we have internal flags already in-use that user space
should not be able to set). But for completeness we have two UAPI
paths into sendpage, sendfile and splice.
In the sendfile case the function do_sendfile() zero's flags,
./fs/read_write.c:
static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos,
size_t count, loff_t max)
{
...
fl = 0;
#if 0
/*
* We need to debate whether we can enable this or not. The
* man page documents EAGAIN return for the output at least,
* and the application is arguably buggy if it doesn't expect
* EAGAIN on a non-blocking file descriptor.
*/
if (in.file->f_flags & O_NONBLOCK)
fl = SPLICE_F_NONBLOCK;
#endif
file_start_write(out.file);
retval = do_splice_direct(in.file, &pos, out.file, &out_pos, count, fl);
}
In the splice case the pipe_to_sendpage "actor" is used which
masks flags with SPLICE_F_MORE.
./fs/splice.c:
static int pipe_to_sendpage(struct pipe_inode_info *pipe,
struct pipe_buffer *buf, struct splice_desc *sd)
{
...
more = (sd->flags & SPLICE_F_MORE) ? MSG_MORE : 0;
...
}
Confirming what we expect that internal flags are in fact internal
to socket side.
Fixes: d3b18ad31f93 ("tls: add bpf support to sk_msg handling")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2018-12-21 03:35:35 +08:00
|
|
|
static inline bool tls_sw_has_ctx_tx(const struct sock *sk)
|
|
|
|
{
|
|
|
|
struct tls_context *ctx = tls_get_ctx(sk);
|
|
|
|
|
|
|
|
if (!ctx)
|
|
|
|
return false;
|
|
|
|
return !!tls_sw_ctx_tx(ctx);
|
|
|
|
}
|
|
|
|
|
2018-07-13 19:33:43 +08:00
|
|
|
static inline struct tls_offload_context_rx *
|
|
|
|
tls_offload_ctx_rx(const struct tls_context *tls_ctx)
|
|
|
|
{
|
|
|
|
return (struct tls_offload_context_rx *)tls_ctx->priv_ctx_rx;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The TLS context is valid until sk_destruct is called */
|
|
|
|
static inline void tls_offload_rx_resync_request(struct sock *sk, __be32 seq)
|
|
|
|
{
|
|
|
|
struct tls_context *tls_ctx = tls_get_ctx(sk);
|
|
|
|
struct tls_offload_context_rx *rx_ctx = tls_offload_ctx_rx(tls_ctx);
|
|
|
|
|
|
|
|
atomic64_set(&rx_ctx->resync_req, ((((uint64_t)seq) << 32) | 1));
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
int tls_proccess_cmsg(struct sock *sk, struct msghdr *msg,
|
|
|
|
unsigned char *record_type);
|
2018-04-01 00:11:52 +08:00
|
|
|
void tls_register_device(struct tls_device *device);
|
|
|
|
void tls_unregister_device(struct tls_device *device);
|
2018-07-13 19:33:43 +08:00
|
|
|
int tls_device_decrypted(struct sock *sk, struct sk_buff *skb);
|
2018-07-13 19:33:40 +08:00
|
|
|
int decrypt_skb(struct sock *sk, struct sk_buff *skb,
|
|
|
|
struct scatterlist *sgout);
|
2017-06-15 02:37:39 +08:00
|
|
|
|
2018-04-30 15:16:16 +08:00
|
|
|
struct sk_buff *tls_validate_xmit_skb(struct sock *sk,
|
|
|
|
struct net_device *dev,
|
|
|
|
struct sk_buff *skb);
|
|
|
|
|
|
|
|
int tls_sw_fallback_init(struct sock *sk,
|
2018-07-13 19:33:39 +08:00
|
|
|
struct tls_offload_context_tx *offload_ctx,
|
2018-04-30 15:16:16 +08:00
|
|
|
struct tls_crypto_info *crypto_info);
|
|
|
|
|
2018-07-13 19:33:43 +08:00
|
|
|
int tls_set_device_offload_rx(struct sock *sk, struct tls_context *ctx);
|
|
|
|
|
|
|
|
void tls_device_offload_cleanup_rx(struct sock *sk);
|
|
|
|
void handle_device_resync(struct sock *sk, u32 seq, u64 rcd_sn);
|
|
|
|
|
2017-06-15 02:37:39 +08:00
|
|
|
#endif /* _TLS_OFFLOAD_H */
|