* Re: [PATCH net-next v2 2/2] bpf: Remove the capability check for cgroup skb eBPF program
From: Daniel Borkmann @ 2017-06-06 16:56 UTC (permalink / raw)
To: Alexei Starovoitov, Chenbo Feng
Cc: netdev, David Miller, Lorenzo Colitti, Chenbo Feng
In-Reply-To: <20170601234235.iwu55crijtxuq5mp@ast-mbp>
On 06/02/2017 01:42 AM, Alexei Starovoitov wrote:
> On Wed, May 31, 2017 at 06:16:00PM -0700, Chenbo Feng wrote:
>> From: Chenbo Feng <fengc@google.com>
>>
>> Currently loading a cgroup skb eBPF program require a CAP_SYS_ADMIN
>> capability while attaching the program to a cgroup only requires the
>> user have CAP_NET_ADMIN privilege. We can escape the capability
>> check when load the program just like socket filter program to make
>> the capability requirement consistent.
>>
>> Change since v1:
>> Change the code style in order to be compliant with checkpatch.pl
>> preference
>>
>> Signed-off-by: Chenbo Feng <fengc@google.com>
>
> as far as I can see they're indeed the same as socket filters, so
> Acked-by: Alexei Starovoitov <ast@kernel.org>
>
> but I don't quite understand how it helps, since as you said
> attaching such unpriv fd to cgroup still requires root.
> Do you have more patches to follow?
Hmm, when we relax this from capable(CAP_SYS_ADMIN) to unprivileged,
then we must at least also zero out the not-yet-initialized memory
for the mac header for egress case in __cgroup_bpf_run_filter_skb().
^ permalink raw reply
* [PATCH v2 net-next 0/4] kernel TLS
From: Dave Watson @ 2017-06-06 16:59 UTC (permalink / raw)
To: Ilya Lesokhin, Aviad Yehezkel, Boris Pismenny, Liran Liss,
Matan Barak, David Miller, netdev, Tom Herbert, herbert,
linux-crypto, Hannes Frederic Sowa, Eric Dumazet
Cc: Alexei Starovoitov, nmav, fridolin.pokorny
This series adds support for kernel TLS encryption over TCP sockets.
A standard TCP socket is converted to a TLS socket using a setsockopt.
Only symmetric crypto is done in the kernel, as well as TLS record
framing. The handshake remains in userspace, and the negotiated
cipher keys/iv are provided to the TCP socket.
We implemented support for this API in OpenSSL 1.1.0, the code is
available at https://github.com/Mellanox/tls-openssl/tree/master
It should work with any TLS library with similar modifications,
a test tool using gnutls is here: https://github.com/Mellanox/tls-af_ktls_tool
Changes from V1:
* EXPORT_SYMBOL GPL in patch 2
* Add link to OpenSSL patch & gnutls example in documentation patch.
* sk_write_pending check was rolled in to wait_for_memory path,
avoids special case and fixes lock inbalance issue.
* Unify flag handling for sendmsg/sendfile
Changes from RFC V2:
* Generic ULP (upper layer protocol) framework instead of TLS specific
setsockopts
* Dropped Mellanox hardware patches, will come as separate series.
Framework will work for both.
RFC V2:
http://www.mail-archive.com/netdev@vger.kernel.org/msg160317.html
Changes from RFC V1:
* Socket based on changing TCP proto_ops instead of crypto framework
* Merged code with Mellanox's hardware tls offload
* Zerocopy sendmsg support added - sendpage/sendfile is no longer
necessary for zerocopy optimization
RFC V1:
http://www.mail-archive.com/netdev@vger.kernel.org/msg88021.html
* Socket based on crypto userspace API framework, required two
sockets in userspace, one encrypted, one unencrypted.
Paper: https://netdevconf.org/1.2/papers/ktls.pdf
Aviad Yehezkel (1):
tcp: export do_tcp_sendpages and tcp_rate_check_app_limited functions
Boris Pismenny (2):
tcp: ULP infrastructure
tls: Documentation
Ilya Lesokhin (1):
tls: kernel TLS support
Documentation/networking/tls.txt | 135 +++++++
MAINTAINERS | 10 +
include/linux/socket.h | 1 +
include/net/inet_connection_sock.h | 4 +
include/net/tcp.h | 27 ++
include/net/tls.h | 222 +++++++++++
include/uapi/linux/tcp.h | 1 +
include/uapi/linux/tls.h | 79 ++++
net/Kconfig | 1 +
net/Makefile | 1 +
net/ipv4/Makefile | 2 +-
net/ipv4/sysctl_net_ipv4.c | 25 ++
net/ipv4/tcp.c | 33 +-
net/ipv4/tcp_ipv4.c | 2 +
net/ipv4/tcp_rate.c | 1 +
net/ipv4/tcp_ulp.c | 134 +++++++
net/tls/Kconfig | 12 +
net/tls/Makefile | 7 +
net/tls/tls_main.c | 485 ++++++++++++++++++++++
net/tls/tls_sw.c | 794 +++++++++++++++++++++++++++++++++++++
20 files changed, 1973 insertions(+), 3 deletions(-)
create mode 100644 Documentation/networking/tls.txt
create mode 100644 include/net/tls.h
create mode 100644 include/uapi/linux/tls.h
create mode 100644 net/ipv4/tcp_ulp.c
create mode 100644 net/tls/Kconfig
create mode 100644 net/tls/Makefile
create mode 100644 net/tls/tls_main.c
create mode 100644 net/tls/tls_sw.c
--
2.9.3
^ permalink raw reply
* [PATCH v2 net-next 1/4] tcp: ULP infrastructure
From: Dave Watson @ 2017-06-06 17:00 UTC (permalink / raw)
To: Ilya Lesokhin, Aviad Yehezkel, Boris Pismenny, Liran Liss,
Matan Barak, David Miller, netdev, Tom Herbert, herbert,
linux-crypto, Hannes Frederic Sowa, Eric Dumazet
Cc: Alexei Starovoitov, nmav, fridolin.pokorny
In-Reply-To: <cover.1496767702.git.davejwatson@fb.com>
Add the infrustructure for attaching Upper Layer Protocols (ULPs) over TCP
sockets. Based on a similar infrastructure in tcp_cong. The idea is that any
ULP can add its own logic by changing the TCP proto_ops structure to its own
methods.
Example usage:
setsockopt(sock, SOL_TCP, TCP_ULP, "tls", sizeof("tls"));
modules will call:
tcp_register_ulp(&tcp_tls_ulp_ops);
to register/unregister their ulp, with an init function and name.
A list of registered ulps will be returned by tcp_get_available_ulp, which is
hooked up to /proc. Example:
$ cat /proc/sys/net/ipv4/tcp_available_ulp
tls
There is currently no functionality to remove or chain ULPs, but
it should be possible to add these in the future if needed.
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Dave Watson <davejwatson@fb.com>
---
include/net/inet_connection_sock.h | 4 ++
include/net/tcp.h | 25 +++++++
include/uapi/linux/tcp.h | 1 +
net/ipv4/Makefile | 2 +-
net/ipv4/sysctl_net_ipv4.c | 25 +++++++
net/ipv4/tcp.c | 28 ++++++++
net/ipv4/tcp_ipv4.c | 2 +
net/ipv4/tcp_ulp.c | 134 +++++++++++++++++++++++++++++++++++++
8 files changed, 220 insertions(+), 1 deletion(-)
create mode 100644 net/ipv4/tcp_ulp.c
diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h
index c7a5779..13e4c89 100644
--- a/include/net/inet_connection_sock.h
+++ b/include/net/inet_connection_sock.h
@@ -75,6 +75,8 @@ struct inet_connection_sock_af_ops {
* @icsk_pmtu_cookie Last pmtu seen by socket
* @icsk_ca_ops Pluggable congestion control hook
* @icsk_af_ops Operations which are AF_INET{4,6} specific
+ * @icsk_ulp_ops Pluggable ULP control hook
+ * @icsk_ulp_data ULP private data
* @icsk_ca_state: Congestion control state
* @icsk_retransmits: Number of unrecovered [RTO] timeouts
* @icsk_pending: Scheduled timer event
@@ -97,6 +99,8 @@ struct inet_connection_sock {
__u32 icsk_pmtu_cookie;
const struct tcp_congestion_ops *icsk_ca_ops;
const struct inet_connection_sock_af_ops *icsk_af_ops;
+ const struct tcp_ulp_ops *icsk_ulp_ops;
+ void *icsk_ulp_data;
unsigned int (*icsk_sync_mss)(struct sock *sk, u32 pmtu);
__u8 icsk_ca_state:6,
icsk_ca_setsockopt:1,
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 82462db..fcc39f8 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1992,4 +1992,29 @@ static inline void tcp_listendrop(const struct sock *sk)
enum hrtimer_restart tcp_pace_kick(struct hrtimer *timer);
+/*
+ * Interface for adding Upper Level Protocols over TCP
+ */
+
+#define TCP_ULP_NAME_MAX 16
+#define TCP_ULP_MAX 128
+#define TCP_ULP_BUF_MAX (TCP_ULP_NAME_MAX*TCP_ULP_MAX)
+
+struct tcp_ulp_ops {
+ struct list_head list;
+
+ /* initialize ulp */
+ int (*init)(struct sock *sk);
+ /* cleanup ulp */
+ void (*release)(struct sock *sk);
+
+ char name[TCP_ULP_NAME_MAX];
+ struct module *owner;
+};
+int tcp_register_ulp(struct tcp_ulp_ops *type);
+void tcp_unregister_ulp(struct tcp_ulp_ops *type);
+int tcp_set_ulp(struct sock *sk, const char *name);
+void tcp_get_available_ulp(char *buf, size_t len);
+void tcp_cleanup_ulp(struct sock *sk);
+
#endif /* _TCP_H */
diff --git a/include/uapi/linux/tcp.h b/include/uapi/linux/tcp.h
index 38a2b07..8204dce 100644
--- a/include/uapi/linux/tcp.h
+++ b/include/uapi/linux/tcp.h
@@ -117,6 +117,7 @@ enum {
#define TCP_SAVED_SYN 28 /* Get SYN headers recorded for connection */
#define TCP_REPAIR_WINDOW 29 /* Get/set window parameters */
#define TCP_FASTOPEN_CONNECT 30 /* Attempt FastOpen with connect */
+#define TCP_ULP 31 /* Attach a ULP to a TCP connection */
struct tcp_repair_opt {
__u32 opt_code;
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index f83de23..afcb435 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -8,7 +8,7 @@ obj-y := route.o inetpeer.o protocol.o \
inet_timewait_sock.o inet_connection_sock.o \
tcp.o tcp_input.o tcp_output.o tcp_timer.o tcp_ipv4.o \
tcp_minisocks.o tcp_cong.o tcp_metrics.o tcp_fastopen.o \
- tcp_rate.o tcp_recovery.o \
+ tcp_rate.o tcp_recovery.o tcp_ulp.o \
tcp_offload.o datagram.o raw.o udp.o udplite.o \
udp_offload.o arp.o icmp.o devinet.o af_inet.o igmp.o \
fib_frontend.o fib_semantics.o fib_trie.o fib_notifier.o \
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 86957e9..6a40837c 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -360,6 +360,25 @@ static int proc_tfo_blackhole_detect_timeout(struct ctl_table *table,
ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
if (write && ret == 0)
tcp_fastopen_active_timeout_reset();
+
+ return ret;
+}
+
+static int proc_tcp_available_ulp(struct ctl_table *ctl,
+ int write,
+ void __user *buffer, size_t *lenp,
+ loff_t *ppos)
+{
+ struct ctl_table tbl = { .maxlen = TCP_ULP_BUF_MAX, };
+ int ret;
+
+ tbl.data = kmalloc(tbl.maxlen, GFP_USER);
+ if (!tbl.data)
+ return -ENOMEM;
+ tcp_get_available_ulp(tbl.data, TCP_ULP_BUF_MAX);
+ ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
+ kfree(tbl.data);
+
return ret;
}
@@ -707,6 +726,12 @@ static struct ctl_table ipv4_table[] = {
.proc_handler = proc_dointvec_ms_jiffies,
},
{
+ .procname = "tcp_available_ulp",
+ .maxlen = TCP_ULP_BUF_MAX,
+ .mode = 0444,
+ .proc_handler = proc_tcp_available_ulp,
+ },
+ {
.procname = "icmp_msgs_per_sec",
.data = &sysctl_icmp_msgs_per_sec,
.maxlen = sizeof(int),
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index f7be94f..0aa72cd 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2461,6 +2461,24 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
release_sock(sk);
return err;
}
+ case TCP_ULP: {
+ char name[TCP_ULP_NAME_MAX];
+
+ if (optlen < 1)
+ return -EINVAL;
+
+ val = strncpy_from_user(name, optval,
+ min_t(long, TCP_ULP_NAME_MAX - 1,
+ optlen));
+ if (val < 0)
+ return -EFAULT;
+ name[val] = 0;
+
+ lock_sock(sk);
+ err = tcp_set_ulp(sk, name);
+ release_sock(sk);
+ return err;
+ }
default:
/* fallthru */
break;
@@ -3017,6 +3035,16 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
return -EFAULT;
return 0;
+ case TCP_ULP:
+ if (get_user(len, optlen))
+ return -EFAULT;
+ len = min_t(unsigned int, len, TCP_ULP_NAME_MAX);
+ if (put_user(len, optlen))
+ return -EFAULT;
+ if (copy_to_user(optval, icsk->icsk_ulp_ops->name, len))
+ return -EFAULT;
+ return 0;
+
case TCP_THIN_LINEAR_TIMEOUTS:
val = tp->thin_lto;
break;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 191b2f7..c2f5538 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1860,6 +1860,8 @@ void tcp_v4_destroy_sock(struct sock *sk)
tcp_cleanup_congestion_control(sk);
+ tcp_cleanup_ulp(sk);
+
/* Cleanup up the write buffer. */
tcp_write_queue_purge(sk);
diff --git a/net/ipv4/tcp_ulp.c b/net/ipv4/tcp_ulp.c
new file mode 100644
index 0000000..e2beb80
--- /dev/null
+++ b/net/ipv4/tcp_ulp.c
@@ -0,0 +1,134 @@
+/*
+ * Pluggable TCP upper layer protocol support.
+ *
+ * Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2016-2017, Dave Watson <davejwatson@fb.com>. All rights reserved.
+ *
+ */
+
+#include<linux/module.h>
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/gfp.h>
+#include <net/tcp.h>
+
+static DEFINE_SPINLOCK(tcp_ulp_list_lock);
+static LIST_HEAD(tcp_ulp_list);
+
+/* Simple linear search, don't expect many entries! */
+static struct tcp_ulp_ops *tcp_ulp_find(const char *name)
+{
+ struct tcp_ulp_ops *e;
+
+ list_for_each_entry_rcu(e, &tcp_ulp_list, list) {
+ if (strcmp(e->name, name) == 0)
+ return e;
+ }
+
+ return NULL;
+}
+
+static const struct tcp_ulp_ops *__tcp_ulp_find_autoload(const char *name)
+{
+ const struct tcp_ulp_ops *ulp = NULL;
+
+ rcu_read_lock();
+ ulp = tcp_ulp_find(name);
+
+#ifdef CONFIG_MODULES
+ if (!ulp && capable(CAP_NET_ADMIN)) {
+ rcu_read_unlock();
+ request_module("%s", name);
+ rcu_read_lock();
+ ulp = tcp_ulp_find(name);
+ }
+#endif
+ if (!ulp || !try_module_get(ulp->owner))
+ ulp = NULL;
+
+ rcu_read_unlock();
+ return ulp;
+}
+
+/* Attach new upper layer protocol to the list
+ * of available protocols.
+ */
+int tcp_register_ulp(struct tcp_ulp_ops *ulp)
+{
+ int ret = 0;
+
+ spin_lock(&tcp_ulp_list_lock);
+ if (tcp_ulp_find(ulp->name)) {
+ pr_notice("%s already registered or non-unique name\n",
+ ulp->name);
+ ret = -EEXIST;
+ } else {
+ list_add_tail_rcu(&ulp->list, &tcp_ulp_list);
+ }
+ spin_unlock(&tcp_ulp_list_lock);
+
+ return ret;
+}
+EXPORT_SYMBOL(tcp_register_ulp);
+
+void tcp_unregister_ulp(struct tcp_ulp_ops *ulp)
+{
+ spin_lock(&tcp_ulp_list_lock);
+ list_del_rcu(&ulp->list);
+ spin_unlock(&tcp_ulp_list_lock);
+
+ synchronize_rcu();
+}
+EXPORT_SYMBOL(tcp_unregister_ulp);
+
+/* Build string with list of available upper layer protocl values */
+void tcp_get_available_ulp(char *buf, size_t maxlen)
+{
+ struct tcp_ulp_ops *ulp_ops;
+ size_t offs = 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(ulp_ops, &tcp_ulp_list, list) {
+ offs += snprintf(buf + offs, maxlen - offs,
+ "%s%s",
+ offs == 0 ? "" : " ", ulp_ops->name);
+ }
+ rcu_read_unlock();
+}
+
+void tcp_cleanup_ulp(struct sock *sk)
+{
+ struct inet_connection_sock *icsk = inet_csk(sk);
+
+ if (!icsk->icsk_ulp_ops)
+ return;
+
+ if (icsk->icsk_ulp_ops->release)
+ icsk->icsk_ulp_ops->release(sk);
+ module_put(icsk->icsk_ulp_ops->owner);
+}
+
+/* Change upper layer protocol for socket */
+int tcp_set_ulp(struct sock *sk, const char *name)
+{
+ struct inet_connection_sock *icsk = inet_csk(sk);
+ const struct tcp_ulp_ops *ulp_ops;
+ int err = 0;
+
+ if (icsk->icsk_ulp_ops)
+ return -EEXIST;
+
+ ulp_ops = __tcp_ulp_find_autoload(name);
+ if (!ulp_ops)
+ err = -ENOENT;
+ else
+ err = ulp_ops->init(sk);
+
+ if (err)
+ goto out;
+
+ icsk->icsk_ulp_ops = ulp_ops;
+ out:
+ return err;
+}
--
2.9.3
^ permalink raw reply related
* [PATCH v2 net-next 2/4] tcp: export do_tcp_sendpages and tcp_rate_check_app_limited functions
From: Dave Watson @ 2017-06-06 17:00 UTC (permalink / raw)
To: Ilya Lesokhin, Aviad Yehezkel, Boris Pismenny, Liran Liss,
Matan Barak, David Miller, netdev, Tom Herbert, herbert,
linux-crypto, Hannes Frederic Sowa, Eric Dumazet
Cc: Alexei Starovoitov, nmav, fridolin.pokorny
In-Reply-To: <cover.1496767702.git.davejwatson@fb.com>
Export do_tcp_sendpages and tcp_rate_check_app_limited, since tls will need to
sendpages while the socket is already locked.
tcp_sendpage is exported, but requires the socket lock to not be held already.
Signed-off-by: Aviad Yehezkel <aviadye@mellanox.com>
Signed-off-by: Ilya Lesokhin <ilyal@mellanox.com>
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Dave Watson <davejwatson@fb.com>
---
include/net/tcp.h | 2 ++
net/ipv4/tcp.c | 5 +++--
net/ipv4/tcp_rate.c | 1 +
3 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index fcc39f8..2b35100 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -353,6 +353,8 @@ int tcp_v4_tw_remember_stamp(struct inet_timewait_sock *tw);
int tcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t size);
int tcp_sendpage(struct sock *sk, struct page *page, int offset, size_t size,
int flags);
+ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset,
+ size_t size, int flags);
void tcp_release_cb(struct sock *sk);
void tcp_wfree(struct sk_buff *skb);
void tcp_write_timer_handler(struct sock *sk);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 0aa72cd..70efada 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -882,8 +882,8 @@ static int tcp_send_mss(struct sock *sk, int *size_goal, int flags)
return mss_now;
}
-static ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset,
- size_t size, int flags)
+ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset,
+ size_t size, int flags)
{
struct tcp_sock *tp = tcp_sk(sk);
int mss_now, size_goal;
@@ -1013,6 +1013,7 @@ static ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset,
}
return sk_stream_error(sk, flags, err);
}
+EXPORT_SYMBOL_GPL(do_tcp_sendpages);
int tcp_sendpage(struct sock *sk, struct page *page, int offset,
size_t size, int flags)
diff --git a/net/ipv4/tcp_rate.c b/net/ipv4/tcp_rate.c
index ad99569..3330a37 100644
--- a/net/ipv4/tcp_rate.c
+++ b/net/ipv4/tcp_rate.c
@@ -185,3 +185,4 @@ void tcp_rate_check_app_limited(struct sock *sk)
tp->app_limited =
(tp->delivered + tcp_packets_in_flight(tp)) ? : 1;
}
+EXPORT_SYMBOL_GPL(tcp_rate_check_app_limited);
--
2.9.3
^ permalink raw reply related
* [PATCH v2 net-next 3/4] tls: kernel TLS support
From: Dave Watson @ 2017-06-06 17:01 UTC (permalink / raw)
To: Ilya Lesokhin, Aviad Yehezkel, Boris Pismenny, Liran Liss,
Matan Barak, David Miller, netdev, Tom Herbert, herbert,
linux-crypto, Hannes Frederic Sowa, Eric Dumazet
Cc: Alexei Starovoitov, nmav, fridolin.pokorny
In-Reply-To: <cover.1496767702.git.davejwatson@fb.com>
Software implementation of transport layer security, implemented using ULP
infrastructure. tcp proto_ops are replaced with tls equivalents of sendmsg and
sendpage.
Only symmetric crypto is done in the kernel, keys are passed by setsockopt
after the handshake is complete. All control messages are supported via CMSG
data - the actual symmetric encryption is the same, just the message type needs
to be passed separately.
For user API, please see Documentation patch.
Pieces that can be shared between hw and sw implementation
are in tls_main.c
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Ilya Lesokhin <ilyal@mellanox.com>
Signed-off-by: Aviad Yehezkel <aviadye@mellanox.com>
Signed-off-by: Dave Watson <davejwatson@fb.com>
---
MAINTAINERS | 10 +
include/linux/socket.h | 1 +
include/net/tls.h | 222 +++++++++++++
include/uapi/linux/tls.h | 79 +++++
net/Kconfig | 1 +
net/Makefile | 1 +
net/tls/Kconfig | 12 +
net/tls/Makefile | 7 +
net/tls/tls_main.c | 485 +++++++++++++++++++++++++++++
net/tls/tls_sw.c | 794 +++++++++++++++++++++++++++++++++++++++++++++++
10 files changed, 1612 insertions(+)
create mode 100644 include/net/tls.h
create mode 100644 include/uapi/linux/tls.h
create mode 100644 net/tls/Kconfig
create mode 100644 net/tls/Makefile
create mode 100644 net/tls/tls_main.c
create mode 100644 net/tls/tls_sw.c
diff --git a/MAINTAINERS b/MAINTAINERS
index 6b7625f..246ddd7 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8973,6 +8973,16 @@ F: net/ipv6/
F: include/net/ip*
F: arch/x86/net/*
+NETWORKING [TLS]
+M: Ilya Lesokhin <ilyal@mellanox.com>
+M: Aviad Yehezkel <aviadye@mellanox.com>
+M: Dave Watson <davejwatson@fb.com>
+L: netdev@vger.kernel.org
+S: Maintained
+F: net/tls/*
+F: include/uapi/linux/tls.h
+F: include/net/tls.h
+
NETWORKING [IPSEC]
M: Steffen Klassert <steffen.klassert@secunet.com>
M: Herbert Xu <herbert@gondor.apana.org.au>
diff --git a/include/linux/socket.h b/include/linux/socket.h
index 0820274..8b13db5 100644
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -334,6 +334,7 @@ struct ucred {
#define SOL_ALG 279
#define SOL_NFC 280
#define SOL_KCM 281
+#define SOL_TLS 282
/* IPX options */
#define IPX_TYPE 1
diff --git a/include/net/tls.h b/include/net/tls.h
new file mode 100644
index 0000000..b20fd2f
--- /dev/null
+++ b/include/net/tls.h
@@ -0,0 +1,222 @@
+/*
+ * 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>
+
+#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
+
+struct tls_sw_context {
+ struct crypto_aead *aead_send;
+
+ /* Sending context */
+ char aad_space[TLS_AAD_SPACE_SIZE];
+
+ unsigned int sg_plaintext_size;
+ int sg_plaintext_num_elem;
+ struct scatterlist sg_plaintext_data[MAX_SKB_FRAGS];
+
+ unsigned int sg_encrypted_size;
+ int sg_encrypted_num_elem;
+ struct scatterlist sg_encrypted_data[MAX_SKB_FRAGS];
+
+ /* AAD | sg_plaintext_data | sg_tag */
+ struct scatterlist sg_aead_in[2];
+ /* AAD | sg_encrypted_data (data contain overhead for hdr&iv&tag) */
+ struct scatterlist sg_aead_out[2];
+};
+
+enum {
+ TLS_PENDING_CLOSED_RECORD
+};
+
+struct tls_context {
+ union {
+ struct tls_crypto_info crypto_send;
+ struct tls12_crypto_info_aes_gcm_128 crypto_send_aes_gcm_128;
+ };
+
+ void *priv_ctx;
+
+ u16 prepend_size;
+ u16 tag_size;
+ u16 overhead_size;
+ u16 iv_size;
+ char *iv;
+ u16 rec_seq_size;
+ char *rec_seq;
+
+ struct scatterlist *partially_sent_record;
+ u16 partially_sent_offset;
+ unsigned long flags;
+
+ u16 pending_open_record_frags;
+ int (*push_pending_record)(struct sock *sk, int flags);
+
+ void (*sk_write_space)(struct sock *sk);
+ void (*sk_destruct)(struct sock *sk);
+
+ 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);
+};
+
+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);
+
+
+int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx);
+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);
+
+void tls_sk_destruct(struct sock *sk, struct tls_context *ctx);
+void tls_icsk_clean_acked(struct sock *sk);
+
+int tls_push_sg(struct sock *sk, struct tls_context *ctx,
+ struct scatterlist *sg, u16 first_offset,
+ int flags);
+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 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;
+}
+
+static inline void tls_err_abort(struct sock *sk)
+{
+ sk->sk_err = -EBADMSG;
+ 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,
+ struct tls_context *ctx)
+{
+ if (tls_bigint_increment(ctx->rec_seq, ctx->rec_seq_size))
+ tls_err_abort(sk);
+ 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)
+{
+ size_t pkt_len, iv_size = ctx->iv_size;
+
+ pkt_len = plaintext_len + iv_size + ctx->tag_size;
+
+ /* 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;
+ buf[1] = TLS_VERSION_MINOR(ctx->crypto_send.version);
+ buf[2] = TLS_VERSION_MAJOR(ctx->crypto_send.version);
+ /* 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,
+ ctx->iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, iv_size);
+}
+
+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;
+}
+
+static inline struct tls_sw_context *tls_sw_ctx(
+ const struct tls_context *tls_ctx)
+{
+ return (struct tls_sw_context *)tls_ctx->priv_ctx;
+}
+
+static inline struct tls_offload_context *tls_offload_ctx(
+ const struct tls_context *tls_ctx)
+{
+ return (struct tls_offload_context *)tls_ctx->priv_ctx;
+}
+
+int tls_proccess_cmsg(struct sock *sk, struct msghdr *msg,
+ unsigned char *record_type);
+
+#endif /* _TLS_OFFLOAD_H */
diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h
new file mode 100644
index 0000000..cc1d21d
--- /dev/null
+++ b/include/uapi/linux/tls.h
@@ -0,0 +1,79 @@
+/*
+ * Copyright (c) 2016-2017, Mellanox Technologies. 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 _UAPI_LINUX_TLS_H
+#define _UAPI_LINUX_TLS_H
+
+#include <linux/types.h>
+#include <asm/byteorder.h>
+#include <linux/socket.h>
+#include <linux/tcp.h>
+#include <net/tcp.h>
+
+/* TLS socket options */
+#define TLS_TX 1 /* Set transmit parameters */
+
+/* Supported versions */
+#define TLS_VERSION_MINOR(ver) ((ver) & 0xFF)
+#define TLS_VERSION_MAJOR(ver) (((ver) >> 8) & 0xFF)
+
+#define TLS_VERSION_NUMBER(id) ((((id##_VERSION_MAJOR) & 0xFF) << 8) | \
+ ((id##_VERSION_MINOR) & 0xFF))
+
+#define TLS_1_2_VERSION_MAJOR 0x3
+#define TLS_1_2_VERSION_MINOR 0x3
+#define TLS_1_2_VERSION TLS_VERSION_NUMBER(TLS_1_2)
+
+/* Supported ciphers */
+#define TLS_CIPHER_AES_GCM_128 51
+#define TLS_CIPHER_AES_GCM_128_IV_SIZE 8
+#define TLS_CIPHER_AES_GCM_128_KEY_SIZE 16
+#define TLS_CIPHER_AES_GCM_128_SALT_SIZE 4
+#define TLS_CIPHER_AES_GCM_128_TAG_SIZE 16
+#define TLS_CIPHER_AES_GCM_128_REC_SEQ_SIZE 8
+
+#define TLS_SET_RECORD_TYPE 1
+
+struct tls_crypto_info {
+ __u16 version;
+ __u16 cipher_type;
+};
+
+struct tls12_crypto_info_aes_gcm_128 {
+ struct tls_crypto_info info;
+ unsigned char iv[TLS_CIPHER_AES_GCM_128_IV_SIZE];
+ unsigned char key[TLS_CIPHER_AES_GCM_128_KEY_SIZE];
+ unsigned char salt[TLS_CIPHER_AES_GCM_128_SALT_SIZE];
+ unsigned char rec_seq[TLS_CIPHER_AES_GCM_128_REC_SEQ_SIZE];
+};
+
+#endif /* _UAPI_LINUX_TLS_H */
diff --git a/net/Kconfig b/net/Kconfig
index 102f781..7d57ef3 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -55,6 +55,7 @@ menu "Networking options"
source "net/packet/Kconfig"
source "net/unix/Kconfig"
+source "net/tls/Kconfig"
source "net/xfrm/Kconfig"
source "net/iucv/Kconfig"
source "net/smc/Kconfig"
diff --git a/net/Makefile b/net/Makefile
index 9086ffb..bed80fa 100644
--- a/net/Makefile
+++ b/net/Makefile
@@ -15,6 +15,7 @@ obj-$(CONFIG_LLC) += llc/
obj-$(CONFIG_NET) += ethernet/ 802/ sched/ netlink/ bpf/
obj-$(CONFIG_NETFILTER) += netfilter/
obj-$(CONFIG_INET) += ipv4/
+obj-$(CONFIG_TLS) += tls/
obj-$(CONFIG_XFRM) += xfrm/
obj-$(CONFIG_UNIX) += unix/
obj-$(CONFIG_NET) += ipv6/
diff --git a/net/tls/Kconfig b/net/tls/Kconfig
new file mode 100644
index 0000000..61e5329
--- /dev/null
+++ b/net/tls/Kconfig
@@ -0,0 +1,12 @@
+#
+# TLS configuration
+#
+config TLS
+ tristate "Transport Layer Security support"
+ depends on NET
+ default m
+ ---help---
+ Enable kernel support for TLS protocol. This allows symmetric
+ encryption handling of the TLS protocol to be done in-kernel.
+
+ If unsure, say M.
diff --git a/net/tls/Makefile b/net/tls/Makefile
new file mode 100644
index 0000000..a930fd1
--- /dev/null
+++ b/net/tls/Makefile
@@ -0,0 +1,7 @@
+#
+# Makefile for the TLS subsystem.
+#
+
+obj-$(CONFIG_TLS) += tls.o
+
+tls-y := tls_main.o tls_sw.o
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
new file mode 100644
index 0000000..82937d4
--- /dev/null
+++ b/net/tls/tls_main.c
@@ -0,0 +1,485 @@
+/*
+ * 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.
+ */
+
+#include <linux/module.h>
+
+#include <net/tcp.h>
+#include <net/inet_common.h>
+#include <linux/highmem.h>
+#include <linux/netdevice.h>
+#include <linux/sched/signal.h>
+
+#include <net/tls.h>
+
+MODULE_AUTHOR("Mellanox Technologies");
+MODULE_DESCRIPTION("Transport Layer Security Support");
+MODULE_LICENSE("Dual BSD/GPL");
+
+static struct proto tls_base_prot;
+static struct proto tls_sw_prot;
+
+static int wait_on_pending_writer(struct sock *sk, long *timeo)
+{
+ int rc = 0;
+ DEFINE_WAIT_FUNC(wait, woken_wake_function);
+
+ add_wait_queue(sk_sleep(sk), &wait);
+ while (1) {
+ if (!*timeo) {
+ rc = -EAGAIN;
+ break;
+ }
+
+ if (signal_pending(current)) {
+ rc = sock_intr_errno(*timeo);
+ break;
+ }
+
+ if (sk_wait_event(sk, timeo, !sk->sk_write_pending, &wait))
+ break;
+ }
+ remove_wait_queue(sk_sleep(sk), &wait);
+ return rc;
+}
+
+int tls_push_sg(struct sock *sk,
+ struct tls_context *ctx,
+ struct scatterlist *sg,
+ u16 first_offset,
+ int flags)
+{
+ int sendpage_flags = flags | MSG_SENDPAGE_NOTLAST;
+ int ret = 0;
+ struct page *p;
+ size_t size;
+ int offset = first_offset;
+
+ size = sg->length - offset;
+ offset += sg->offset;
+
+ while (1) {
+ if (sg_is_last(sg))
+ sendpage_flags = flags;
+
+ /* is sending application-limited? */
+ tcp_rate_check_app_limited(sk);
+ p = sg_page(sg);
+retry:
+ ret = do_tcp_sendpages(sk, p, offset, size, sendpage_flags);
+
+ if (ret != size) {
+ if (ret > 0) {
+ offset += ret;
+ size -= ret;
+ goto retry;
+ }
+
+ offset -= sg->offset;
+ ctx->partially_sent_offset = offset;
+ ctx->partially_sent_record = (void *)sg;
+ return ret;
+ }
+
+ put_page(p);
+ sk_mem_uncharge(sk, sg->length);
+ sg = sg_next(sg);
+ if (!sg)
+ break;
+
+ offset = sg->offset;
+ size = sg->length;
+ }
+
+ clear_bit(TLS_PENDING_CLOSED_RECORD, &ctx->flags);
+
+ return 0;
+}
+
+static int tls_handle_open_record(struct sock *sk, int flags)
+{
+ struct tls_context *ctx = tls_get_ctx(sk);
+
+ if (tls_is_pending_open_record(ctx))
+ return ctx->push_pending_record(sk, flags);
+
+ return 0;
+}
+
+int tls_proccess_cmsg(struct sock *sk, struct msghdr *msg,
+ unsigned char *record_type)
+{
+ struct cmsghdr *cmsg;
+ int rc = -EINVAL;
+
+ for_each_cmsghdr(cmsg, msg) {
+ if (!CMSG_OK(msg, cmsg))
+ return -EINVAL;
+ if (cmsg->cmsg_level != SOL_TLS)
+ continue;
+
+ switch (cmsg->cmsg_type) {
+ case TLS_SET_RECORD_TYPE:
+ if (cmsg->cmsg_len < CMSG_LEN(sizeof(*record_type)))
+ return -EINVAL;
+
+ if (msg->msg_flags & MSG_MORE)
+ return -EINVAL;
+
+ rc = tls_handle_open_record(sk, msg->msg_flags);
+ if (rc)
+ return rc;
+
+ *record_type = *(unsigned char *)CMSG_DATA(cmsg);
+ rc = 0;
+ break;
+ default:
+ return -EINVAL;
+ }
+ }
+
+ return rc;
+}
+
+int tls_push_pending_closed_record(struct sock *sk, struct tls_context *ctx,
+ int flags, long *timeo)
+{
+ struct scatterlist *sg;
+ u16 offset;
+ int rc;
+
+ if (unlikely(sk->sk_write_pending)) {
+ rc = wait_on_pending_writer(sk, timeo);
+ if (rc)
+ return rc;
+ }
+
+ if (!tls_is_partially_sent_record(ctx))
+ return ctx->push_pending_record(sk, flags);
+
+ sg = ctx->partially_sent_record;
+ offset = ctx->partially_sent_offset;
+
+ ctx->partially_sent_record = NULL;
+ return tls_push_sg(sk, ctx, sg, offset, flags);
+}
+
+static void tls_write_space(struct sock *sk)
+{
+ struct tls_context *ctx = tls_get_ctx(sk);
+
+ if (!sk->sk_write_pending && tls_is_pending_closed_record(ctx)) {
+ gfp_t sk_allocation = sk->sk_allocation;
+ int rc;
+ long timeo = 0;
+
+ sk->sk_allocation = GFP_ATOMIC;
+ rc = tls_push_pending_closed_record(sk, ctx,
+ MSG_DONTWAIT |
+ MSG_NOSIGNAL,
+ &timeo);
+ sk->sk_allocation = sk_allocation;
+
+ if (rc < 0)
+ return;
+ }
+
+ ctx->sk_write_space(sk);
+}
+
+static void tls_sk_proto_close(struct sock *sk, long timeout)
+{
+ struct tls_context *ctx = tls_get_ctx(sk);
+
+ lock_sock(sk);
+ if (tls_is_pending_closed_record(ctx)) {
+ long timeo = sock_sndtimeo(sk, 0);
+
+ tls_push_pending_closed_record(sk, ctx, 0, &timeo);
+ }
+ tls_handle_open_record(sk, 0);
+ release_sock(sk);
+ ctx->sk_proto_close(sk, timeout);
+}
+
+static int do_tls_getsockopt_tx(struct sock *sk, char __user *optval,
+ int __user *optlen)
+{
+ int rc = 0;
+ struct tls_context *ctx = tls_get_ctx(sk);
+ struct tls_crypto_info *crypto_info;
+ int len;
+
+ if (get_user(len, optlen))
+ return -EFAULT;
+
+ if (!optval || (len < sizeof(*crypto_info))) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ if (!ctx) {
+ rc = -EBUSY;
+ goto out;
+ }
+
+ /* get user crypto info */
+ crypto_info = &ctx->crypto_send;
+
+ if (!TLS_CRYPTO_INFO_READY(crypto_info)) {
+ rc = -EBUSY;
+ goto out;
+ }
+
+ if (len == sizeof(crypto_info)) {
+ rc = copy_to_user(optval, crypto_info, sizeof(*crypto_info));
+ goto out;
+ }
+
+ switch (crypto_info->cipher_type) {
+ case TLS_CIPHER_AES_GCM_128: {
+ struct tls12_crypto_info_aes_gcm_128 *
+ crypto_info_aes_gcm_128 =
+ container_of(crypto_info,
+ struct tls12_crypto_info_aes_gcm_128,
+ info);
+
+ if (len != sizeof(*crypto_info_aes_gcm_128)) {
+ rc = -EINVAL;
+ goto out;
+ }
+ lock_sock(sk);
+ memcpy(crypto_info_aes_gcm_128->iv, ctx->iv,
+ TLS_CIPHER_AES_GCM_128_IV_SIZE);
+ release_sock(sk);
+ rc = copy_to_user(optval,
+ crypto_info_aes_gcm_128,
+ sizeof(*crypto_info_aes_gcm_128));
+ break;
+ }
+ default:
+ rc = -EINVAL;
+ }
+
+out:
+ return rc;
+}
+
+static int do_tls_getsockopt(struct sock *sk, int optname,
+ char __user *optval, int __user *optlen)
+{
+ int rc = 0;
+
+ switch (optname) {
+ case TLS_TX:
+ rc = do_tls_getsockopt_tx(sk, optval, optlen);
+ break;
+ default:
+ rc = -ENOPROTOOPT;
+ break;
+ }
+ return rc;
+}
+
+static int tls_getsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, int __user *optlen)
+{
+ struct tls_context *ctx = tls_get_ctx(sk);
+
+ if (level != SOL_TLS)
+ return ctx->getsockopt(sk, level, optname, optval, optlen);
+
+ return do_tls_getsockopt(sk, optname, optval, optlen);
+}
+
+static int do_tls_setsockopt_tx(struct sock *sk, char __user *optval,
+ unsigned int optlen)
+{
+ struct tls_crypto_info *crypto_info, tmp_crypto_info;
+ struct tls_context *ctx = tls_get_ctx(sk);
+ struct proto *prot = NULL;
+ int rc = 0;
+
+ if (!optval || (optlen < sizeof(*crypto_info))) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ rc = copy_from_user(&tmp_crypto_info, optval, sizeof(*crypto_info));
+ if (rc) {
+ rc = -EFAULT;
+ goto out;
+ }
+
+ /* check version */
+ if (tmp_crypto_info.version != TLS_1_2_VERSION) {
+ rc = -ENOTSUPP;
+ goto out;
+ }
+
+ /* get user crypto info */
+ crypto_info = &ctx->crypto_send;
+
+ /* Currently we don't support set crypto info more than one time */
+ if (TLS_CRYPTO_INFO_READY(crypto_info))
+ goto out;
+
+ switch (tmp_crypto_info.cipher_type) {
+ case TLS_CIPHER_AES_GCM_128: {
+ if (optlen != sizeof(struct tls12_crypto_info_aes_gcm_128)) {
+ rc = -EINVAL;
+ goto out;
+ }
+ rc = copy_from_user(
+ crypto_info,
+ optval,
+ sizeof(struct tls12_crypto_info_aes_gcm_128));
+
+ if (rc) {
+ rc = -EFAULT;
+ goto err_crypto_info;
+ }
+ break;
+ }
+ default:
+ rc = -EINVAL;
+ goto out;
+ }
+
+ ctx->sk_write_space = sk->sk_write_space;
+ sk->sk_write_space = tls_write_space;
+
+ ctx->sk_destruct = sk->sk_destruct;
+
+ ctx->sk_proto_close = sk->sk_prot->close;
+
+ /* currently SW is default, we will have ethtool in future */
+ rc = tls_set_sw_offload(sk, ctx);
+ prot = &tls_sw_prot;
+ if (rc)
+ goto err_crypto_info;
+
+ sk->sk_prot = prot;
+ goto out;
+
+err_crypto_info:
+ memset(crypto_info, 0, sizeof(*crypto_info));
+out:
+ return rc;
+}
+
+static int do_tls_setsockopt(struct sock *sk, int optname,
+ char __user *optval, unsigned int optlen)
+{
+ int rc = 0;
+
+ switch (optname) {
+ case TLS_TX:
+ lock_sock(sk);
+ rc = do_tls_setsockopt_tx(sk, optval, optlen);
+ release_sock(sk);
+ break;
+ default:
+ rc = -ENOPROTOOPT;
+ break;
+ }
+ return rc;
+}
+
+static int tls_setsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, unsigned int optlen)
+{
+ struct tls_context *ctx = tls_get_ctx(sk);
+
+ if (level != SOL_TLS)
+ return ctx->setsockopt(sk, level, optname, optval, optlen);
+
+ return do_tls_setsockopt(sk, optname, optval, optlen);
+}
+
+void tls_sk_destruct(struct sock *sk, struct tls_context *ctx)
+{
+ ctx->sk_destruct(sk);
+ kfree(ctx->rec_seq);
+ kfree(ctx->iv);
+ kfree(ctx);
+}
+
+static int tls_init(struct sock *sk)
+{
+ struct inet_connection_sock *icsk = inet_csk(sk);
+ struct tls_context *ctx;
+ int rc = 0;
+
+ /* allocate tls context */
+ ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+ if (!ctx) {
+ rc = -ENOMEM;
+ goto out;
+ }
+ icsk->icsk_ulp_data = ctx;
+ ctx->setsockopt = sk->sk_prot->setsockopt;
+ ctx->getsockopt = sk->sk_prot->getsockopt;
+ sk->sk_prot = &tls_base_prot;
+out:
+ return rc;
+}
+
+static struct tcp_ulp_ops tcp_tls_ulp_ops __read_mostly = {
+ .name = "tls",
+ .owner = THIS_MODULE,
+ .init = tls_init,
+};
+
+static int __init tls_register(void)
+{
+ tls_base_prot = tcp_prot;
+ tls_base_prot.setsockopt = tls_setsockopt;
+ tls_base_prot.getsockopt = tls_getsockopt;
+
+ tls_sw_prot = tls_base_prot;
+ tls_sw_prot.sendmsg = tls_sw_sendmsg;
+ tls_sw_prot.sendpage = tls_sw_sendpage;
+ tls_sw_prot.close = tls_sk_proto_close;
+
+ tcp_register_ulp(&tcp_tls_ulp_ops);
+
+ return 0;
+}
+
+static void __exit tls_unregister(void)
+{
+ tcp_unregister_ulp(&tcp_tls_ulp_ops);
+}
+
+module_init(tls_register);
+module_exit(tls_unregister);
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
new file mode 100644
index 0000000..86802f4
--- /dev/null
+++ b/net/tls/tls_sw.c
@@ -0,0 +1,794 @@
+/*
+ * Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2016-2017, Dave Watson <davejwatson@fb.com>. All rights reserved.
+ * Copyright (c) 2016-2017, Lance Chao <lancerchao@fb.com>. All rights reserved.
+ * Copyright (c) 2016, Fridolin Pokorny <fridolin.pokorny@gmail.com>. All rights reserved.
+ * Copyright (c) 2016, Nikos Mavrogiannopoulos <nmav@gnults.org>. 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.
+ */
+
+#include <linux/module.h>
+#include <crypto/aead.h>
+
+#include <net/tls.h>
+
+static inline void tls_make_aad(int recv,
+ 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;
+}
+
+static void trim_sg(struct sock *sk, struct scatterlist *sg,
+ int *sg_num_elem, unsigned int *sg_size, int target_size)
+{
+ int i = *sg_num_elem - 1;
+ int trim = *sg_size - target_size;
+
+ if (trim <= 0) {
+ WARN_ON(trim < 0);
+ return;
+ }
+
+ *sg_size = target_size;
+ while (trim >= sg[i].length) {
+ trim -= sg[i].length;
+ sk_mem_uncharge(sk, sg[i].length);
+ put_page(sg_page(&sg[i]));
+ i--;
+
+ if (i < 0)
+ goto out;
+ }
+
+ sg[i].length -= trim;
+ sk_mem_uncharge(sk, trim);
+
+out:
+ *sg_num_elem = i + 1;
+}
+
+static void trim_both_sgl(struct sock *sk, int target_size)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+
+ trim_sg(sk, ctx->sg_plaintext_data,
+ &ctx->sg_plaintext_num_elem,
+ &ctx->sg_plaintext_size,
+ target_size);
+
+ if (target_size > 0)
+ target_size += tls_ctx->overhead_size;
+
+ trim_sg(sk, ctx->sg_encrypted_data,
+ &ctx->sg_encrypted_num_elem,
+ &ctx->sg_encrypted_size,
+ target_size);
+}
+
+static int alloc_sg(struct sock *sk, int len, struct scatterlist *sg,
+ int *sg_num_elem, unsigned int *sg_size,
+ int first_coalesce)
+{
+ struct page_frag *pfrag;
+ unsigned int size = *sg_size;
+ int num_elem = *sg_num_elem, use = 0, rc = 0;
+ struct scatterlist *sge;
+ unsigned int orig_offset;
+
+ len -= size;
+ pfrag = sk_page_frag(sk);
+
+ while (len > 0) {
+ if (!sk_page_frag_refill(sk, pfrag)) {
+ rc = -ENOMEM;
+ goto out;
+ }
+
+ use = min_t(int, len, pfrag->size - pfrag->offset);
+
+ if (!sk_wmem_schedule(sk, use)) {
+ rc = -ENOMEM;
+ goto out;
+ }
+
+ sk_mem_charge(sk, use);
+ size += use;
+ orig_offset = pfrag->offset;
+ pfrag->offset += use;
+
+ sge = sg + num_elem - 1;
+ if (num_elem > first_coalesce && sg_page(sg) == pfrag->page &&
+ sg->offset + sg->length == orig_offset) {
+ sg->length += use;
+ } else {
+ sge++;
+ sg_unmark_end(sge);
+ sg_set_page(sge, pfrag->page, use, orig_offset);
+ get_page(pfrag->page);
+ ++num_elem;
+ if (num_elem == MAX_SKB_FRAGS) {
+ rc = -ENOSPC;
+ break;
+ }
+ }
+
+ len -= use;
+ }
+ goto out;
+
+out:
+ *sg_size = size;
+ *sg_num_elem = num_elem;
+ return rc;
+}
+
+static int alloc_encrypted_sg(struct sock *sk, int len)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ int rc = 0;
+
+ rc = alloc_sg(sk, len, ctx->sg_encrypted_data,
+ &ctx->sg_encrypted_num_elem, &ctx->sg_encrypted_size, 0);
+
+ return rc;
+}
+
+static int alloc_plaintext_sg(struct sock *sk, int len)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ int rc = 0;
+
+ rc = alloc_sg(sk, len, ctx->sg_plaintext_data,
+ &ctx->sg_plaintext_num_elem, &ctx->sg_plaintext_size,
+ tls_ctx->pending_open_record_frags);
+
+ return rc;
+}
+
+static void free_sg(struct sock *sk, struct scatterlist *sg,
+ int *sg_num_elem, unsigned int *sg_size)
+{
+ int i, n = *sg_num_elem;
+
+ for (i = 0; i < n; ++i) {
+ sk_mem_uncharge(sk, sg[i].length);
+ put_page(sg_page(&sg[i]));
+ }
+ *sg_num_elem = 0;
+ *sg_size = 0;
+}
+
+static void tls_free_both_sg(struct sock *sk)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+
+ free_sg(sk, ctx->sg_encrypted_data, &ctx->sg_encrypted_num_elem,
+ &ctx->sg_encrypted_size);
+
+ free_sg(sk, ctx->sg_plaintext_data, &ctx->sg_plaintext_num_elem,
+ &ctx->sg_plaintext_size);
+}
+
+static int tls_do_encryption(struct tls_context *tls_ctx,
+ struct tls_sw_context *ctx, size_t data_len,
+ gfp_t flags)
+{
+ unsigned int req_size = sizeof(struct aead_request) +
+ crypto_aead_reqsize(ctx->aead_send);
+ struct aead_request *aead_req;
+ int rc;
+
+ aead_req = kmalloc(req_size, flags);
+ if (!aead_req)
+ return -ENOMEM;
+
+ ctx->sg_encrypted_data[0].offset += tls_ctx->prepend_size;
+ ctx->sg_encrypted_data[0].length -= tls_ctx->prepend_size;
+
+ aead_request_set_tfm(aead_req, ctx->aead_send);
+ aead_request_set_ad(aead_req, TLS_AAD_SPACE_SIZE);
+ aead_request_set_crypt(aead_req, ctx->sg_aead_in, ctx->sg_aead_out,
+ data_len, tls_ctx->iv);
+ rc = crypto_aead_encrypt(aead_req);
+
+ ctx->sg_encrypted_data[0].offset -= tls_ctx->prepend_size;
+ ctx->sg_encrypted_data[0].length += tls_ctx->prepend_size;
+
+ kfree(aead_req);
+ return rc;
+}
+
+static int tls_push_record(struct sock *sk, int flags,
+ unsigned char record_type)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ int rc;
+
+ sg_mark_end(ctx->sg_plaintext_data + ctx->sg_plaintext_num_elem - 1);
+ sg_mark_end(ctx->sg_encrypted_data + ctx->sg_encrypted_num_elem - 1);
+
+ tls_make_aad(0, ctx->aad_space, ctx->sg_plaintext_size,
+ tls_ctx->rec_seq, tls_ctx->rec_seq_size,
+ record_type);
+
+ tls_fill_prepend(tls_ctx,
+ page_address(sg_page(&ctx->sg_encrypted_data[0])) +
+ ctx->sg_encrypted_data[0].offset,
+ ctx->sg_plaintext_size, record_type);
+
+ tls_ctx->pending_open_record_frags = 0;
+ set_bit(TLS_PENDING_CLOSED_RECORD, &tls_ctx->flags);
+
+ rc = tls_do_encryption(tls_ctx, ctx, ctx->sg_plaintext_size,
+ sk->sk_allocation);
+ if (rc < 0) {
+ /* If we are called from write_space and
+ * we fail, we need to set this SOCK_NOSPACE
+ * to trigger another write_space in the future.
+ */
+ set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
+ return rc;
+ }
+
+ free_sg(sk, ctx->sg_plaintext_data, &ctx->sg_plaintext_num_elem,
+ &ctx->sg_plaintext_size);
+
+ ctx->sg_encrypted_num_elem = 0;
+ ctx->sg_encrypted_size = 0;
+
+ /* Only pass through MSG_DONTWAIT and MSG_NOSIGNAL flags */
+ rc = tls_push_sg(sk, tls_ctx, ctx->sg_encrypted_data, 0, flags);
+ if (rc < 0 && rc != -EAGAIN)
+ tls_err_abort(sk);
+
+ tls_advance_record_sn(sk, tls_ctx);
+ return rc;
+}
+
+static int tls_sw_push_pending_record(struct sock *sk, int flags)
+{
+ return tls_push_record(sk, flags, TLS_RECORD_TYPE_DATA);
+}
+
+static int zerocopy_from_iter(struct sock *sk, struct iov_iter *from,
+ int length)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ struct page *pages[MAX_SKB_FRAGS];
+
+ size_t offset;
+ ssize_t copied, use;
+ int i = 0;
+ unsigned int size = ctx->sg_plaintext_size;
+ int num_elem = ctx->sg_plaintext_num_elem;
+ int rc = 0;
+ int maxpages;
+
+ while (length > 0) {
+ i = 0;
+ maxpages = ARRAY_SIZE(ctx->sg_plaintext_data) - num_elem;
+ if (maxpages == 0) {
+ rc = -EFAULT;
+ goto out;
+ }
+ copied = iov_iter_get_pages(from, pages,
+ length,
+ maxpages, &offset);
+ if (copied <= 0) {
+ rc = -EFAULT;
+ goto out;
+ }
+
+ iov_iter_advance(from, copied);
+
+ length -= copied;
+ size += copied;
+ while (copied) {
+ use = min_t(int, copied, PAGE_SIZE - offset);
+
+ sg_set_page(&ctx->sg_plaintext_data[num_elem],
+ pages[i], use, offset);
+ sg_unmark_end(&ctx->sg_plaintext_data[num_elem]);
+ sk_mem_charge(sk, use);
+
+ offset = 0;
+ copied -= use;
+
+ ++i;
+ ++num_elem;
+ }
+ }
+
+out:
+ ctx->sg_plaintext_size = size;
+ ctx->sg_plaintext_num_elem = num_elem;
+ return rc;
+}
+
+static int memcopy_from_iter(struct sock *sk, struct iov_iter *from,
+ int bytes)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ struct scatterlist *sg = ctx->sg_plaintext_data;
+ int copy, i, rc = 0;
+
+ for (i = tls_ctx->pending_open_record_frags;
+ i < ctx->sg_plaintext_num_elem; ++i) {
+ copy = sg[i].length;
+ if (copy_from_iter(
+ page_address(sg_page(&sg[i])) + sg[i].offset,
+ copy, from) != copy) {
+ rc = -EFAULT;
+ goto out;
+ }
+ bytes -= copy;
+
+ ++tls_ctx->pending_open_record_frags;
+
+ if (!bytes)
+ break;
+ }
+
+out:
+ return rc;
+}
+
+int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ int ret = 0;
+ int required_size;
+ long timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
+ bool eor = !(msg->msg_flags & MSG_MORE);
+ size_t try_to_copy, copied = 0;
+ unsigned char record_type = TLS_RECORD_TYPE_DATA;
+ int record_room;
+ bool full_record;
+ int orig_size;
+
+ if (msg->msg_flags & ~(MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL))
+ return -ENOTSUPP;
+
+ lock_sock(sk);
+
+ if (tls_is_pending_closed_record(tls_ctx)) {
+ ret = tls_push_pending_closed_record(sk, tls_ctx,
+ msg->msg_flags,
+ &timeo);
+ if (ret < 0)
+ goto send_end;
+ }
+
+ if (unlikely(msg->msg_controllen)) {
+ ret = tls_proccess_cmsg(sk, msg, &record_type);
+ if (ret)
+ goto send_end;
+ }
+
+ while (msg_data_left(msg)) {
+ if (sk->sk_err) {
+ ret = sk->sk_err;
+ goto send_end;
+ }
+
+ orig_size = ctx->sg_plaintext_size;
+ full_record = false;
+ try_to_copy = msg_data_left(msg);
+ record_room = TLS_MAX_PAYLOAD_SIZE - ctx->sg_plaintext_size;
+ if (try_to_copy >= record_room) {
+ try_to_copy = record_room;
+ full_record = true;
+ }
+
+ required_size = ctx->sg_plaintext_size + try_to_copy +
+ tls_ctx->overhead_size;
+
+ if (!sk_stream_memory_free(sk))
+ goto wait_for_sndbuf;
+alloc_encrypted:
+ ret = alloc_encrypted_sg(sk, required_size);
+ if (ret) {
+ if (ret != -ENOSPC)
+ goto wait_for_memory;
+
+ /* Adjust try_to_copy according to the amount that was
+ * actually allocated. The difference is due
+ * to max sg elements limit
+ */
+ try_to_copy -= required_size - ctx->sg_encrypted_size;
+ full_record = true;
+ }
+
+ if (full_record || eor) {
+ ret = zerocopy_from_iter(sk, &msg->msg_iter,
+ try_to_copy);
+ if (ret)
+ goto fallback_to_reg_send;
+
+ copied += try_to_copy;
+ ret = tls_push_record(sk, msg->msg_flags, record_type);
+ if (!ret)
+ continue;
+ if (ret == -EAGAIN)
+ goto send_end;
+
+ copied -= try_to_copy;
+fallback_to_reg_send:
+ iov_iter_revert(&msg->msg_iter,
+ ctx->sg_plaintext_size - orig_size);
+ trim_sg(sk, ctx->sg_plaintext_data,
+ &ctx->sg_plaintext_num_elem,
+ &ctx->sg_plaintext_size,
+ orig_size);
+ }
+
+ required_size = ctx->sg_plaintext_size + try_to_copy;
+alloc_plaintext:
+ ret = alloc_plaintext_sg(sk, required_size);
+ if (ret) {
+ if (ret != -ENOSPC)
+ goto wait_for_memory;
+
+ /* Adjust try_to_copy according to the amount that was
+ * actually allocated. The difference is due
+ * to max sg elements limit
+ */
+ try_to_copy -= required_size - ctx->sg_plaintext_size;
+ full_record = true;
+
+ trim_sg(sk, ctx->sg_encrypted_data,
+ &ctx->sg_encrypted_num_elem,
+ &ctx->sg_encrypted_size,
+ ctx->sg_plaintext_size +
+ tls_ctx->overhead_size);
+ }
+
+ ret = memcopy_from_iter(sk, &msg->msg_iter, try_to_copy);
+ if (ret)
+ goto trim_sgl;
+
+ copied += try_to_copy;
+ if (full_record || eor) {
+push_record:
+ ret = tls_push_record(sk, msg->msg_flags, record_type);
+ if (ret) {
+ if (ret == -ENOMEM)
+ goto wait_for_memory;
+
+ goto send_end;
+ }
+ }
+
+ continue;
+
+wait_for_sndbuf:
+ set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
+wait_for_memory:
+ ret = sk_stream_wait_memory(sk, &timeo);
+ if (ret) {
+trim_sgl:
+ trim_both_sgl(sk, orig_size);
+ goto send_end;
+ }
+
+ if (tls_is_pending_closed_record(tls_ctx))
+ goto push_record;
+
+ if (ctx->sg_encrypted_size < required_size)
+ goto alloc_encrypted;
+
+ goto alloc_plaintext;
+ }
+
+send_end:
+ ret = sk_stream_error(sk, msg->msg_flags, ret);
+
+ release_sock(sk);
+ return copied ? copied : ret;
+}
+
+int tls_sw_sendpage(struct sock *sk, struct page *page,
+ int offset, size_t size, int flags)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+ int ret = 0;
+ long timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT);
+ bool eor;
+ size_t orig_size = size;
+ unsigned char record_type = TLS_RECORD_TYPE_DATA;
+ struct scatterlist *sg;
+ bool full_record;
+ int record_room;
+
+ if (flags & ~(MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL |
+ MSG_SENDPAGE_NOTLAST))
+ return -ENOTSUPP;
+
+ /* No MSG_EOR from splice, only look at MSG_MORE */
+ eor = !(flags & (MSG_MORE | MSG_SENDPAGE_NOTLAST));
+
+ lock_sock(sk);
+
+ sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk);
+
+ if (tls_is_pending_closed_record(tls_ctx)) {
+ ret = tls_push_pending_closed_record(sk, tls_ctx, flags,
+ &timeo);
+ if (ret < 0)
+ goto sendpage_end;
+ }
+
+ /* Call the sk_stream functions to manage the sndbuf mem. */
+ while (size > 0) {
+ size_t copy, required_size;
+
+ if (sk->sk_err) {
+ ret = sk->sk_err;
+ goto sendpage_end;
+ }
+
+ full_record = false;
+ record_room = TLS_MAX_PAYLOAD_SIZE - ctx->sg_plaintext_size;
+ copy = size;
+ if (copy >= record_room) {
+ copy = record_room;
+ full_record = true;
+ }
+ required_size = ctx->sg_plaintext_size + copy +
+ tls_ctx->overhead_size;
+
+ if (!sk_stream_memory_free(sk))
+ goto wait_for_sndbuf;
+alloc_payload:
+ ret = alloc_encrypted_sg(sk, required_size);
+ if (ret) {
+ if (ret != -ENOSPC)
+ goto wait_for_memory;
+
+ /* Adjust copy according to the amount that was
+ * actually allocated. The difference is due
+ * to max sg elements limit
+ */
+ copy -= required_size - ctx->sg_plaintext_size;
+ full_record = true;
+ }
+
+ get_page(page);
+ sg = ctx->sg_plaintext_data + ctx->sg_plaintext_num_elem;
+ sg_set_page(sg, page, copy, offset);
+ ctx->sg_plaintext_num_elem++;
+
+ sk_mem_charge(sk, copy);
+ offset += copy;
+ size -= copy;
+ ctx->sg_plaintext_size += copy;
+ tls_ctx->pending_open_record_frags = ctx->sg_plaintext_num_elem;
+
+ if (full_record || eor ||
+ ctx->sg_plaintext_num_elem ==
+ ARRAY_SIZE(ctx->sg_plaintext_data)) {
+push_record:
+ ret = tls_push_record(sk, flags, record_type);
+ if (ret) {
+ if (ret == -ENOMEM)
+ goto wait_for_memory;
+
+ goto sendpage_end;
+ }
+ }
+ continue;
+wait_for_sndbuf:
+ set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
+wait_for_memory:
+ ret = sk_stream_wait_memory(sk, &timeo);
+ if (ret) {
+ trim_both_sgl(sk, ctx->sg_plaintext_size);
+ goto sendpage_end;
+ }
+
+ if (tls_is_pending_closed_record(tls_ctx))
+ goto push_record;
+
+ goto alloc_payload;
+ }
+
+sendpage_end:
+ if (orig_size > size)
+ ret = orig_size - size;
+ else
+ ret = sk_stream_error(sk, flags, ret);
+
+ release_sock(sk);
+ return ret;
+}
+
+void tls_sw_sk_destruct(struct sock *sk)
+{
+ struct tls_context *tls_ctx = tls_get_ctx(sk);
+ struct tls_sw_context *ctx = tls_sw_ctx(tls_ctx);
+
+ if (ctx->aead_send)
+ crypto_free_aead(ctx->aead_send);
+
+ tls_free_both_sg(sk);
+ if (tls_ctx->partially_sent_record) {
+ struct scatterlist *sg = tls_ctx->partially_sent_record;
+
+ while (1) {
+ put_page(sg_page(sg));
+ sk_mem_uncharge(sk, sg->length);
+
+ if (sg_is_last(sg))
+ break;
+ sg++;
+ }
+ }
+
+ kfree(ctx);
+ tls_sk_destruct(sk, tls_ctx);
+}
+
+int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx)
+{
+ char keyval[TLS_CIPHER_AES_GCM_128_KEY_SIZE];
+ struct tls_crypto_info *crypto_info;
+ struct tls12_crypto_info_aes_gcm_128 *gcm_128_info;
+ struct tls_sw_context *sw_ctx;
+ u16 nonce_size, tag_size, iv_size, rec_seq_size;
+ char *iv, *rec_seq;
+ int rc = 0;
+
+ if (!ctx) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ if (ctx->priv_ctx) {
+ rc = -EEXIST;
+ goto out;
+ }
+
+ sw_ctx = kzalloc(sizeof(*sw_ctx), GFP_KERNEL);
+ if (!sw_ctx) {
+ rc = -ENOMEM;
+ goto out;
+ }
+
+ ctx->priv_ctx = (struct tls_offload_context *)sw_ctx;
+
+ crypto_info = &ctx->crypto_send;
+ switch (crypto_info->cipher_type) {
+ case TLS_CIPHER_AES_GCM_128: {
+ nonce_size = TLS_CIPHER_AES_GCM_128_IV_SIZE;
+ tag_size = TLS_CIPHER_AES_GCM_128_TAG_SIZE;
+ iv_size = TLS_CIPHER_AES_GCM_128_IV_SIZE;
+ iv = ((struct tls12_crypto_info_aes_gcm_128 *)crypto_info)->iv;
+ rec_seq_size = TLS_CIPHER_AES_GCM_128_REC_SEQ_SIZE;
+ rec_seq =
+ ((struct tls12_crypto_info_aes_gcm_128 *)crypto_info)->rec_seq;
+ gcm_128_info =
+ (struct tls12_crypto_info_aes_gcm_128 *)crypto_info;
+ break;
+ }
+ default:
+ rc = -EINVAL;
+ goto out;
+ }
+
+ ctx->prepend_size = TLS_HEADER_SIZE + nonce_size;
+ ctx->tag_size = tag_size;
+ ctx->overhead_size = ctx->prepend_size + ctx->tag_size;
+ ctx->iv_size = iv_size;
+ ctx->iv = kmalloc(iv_size + TLS_CIPHER_AES_GCM_128_SALT_SIZE,
+ GFP_KERNEL);
+ if (!ctx->iv) {
+ rc = -ENOMEM;
+ goto out;
+ }
+ memcpy(ctx->iv, gcm_128_info->salt, TLS_CIPHER_AES_GCM_128_SALT_SIZE);
+ memcpy(ctx->iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, iv, iv_size);
+ ctx->rec_seq_size = rec_seq_size;
+ ctx->rec_seq = kmalloc(rec_seq_size, GFP_KERNEL);
+ if (!ctx->rec_seq) {
+ rc = -ENOMEM;
+ goto free_iv;
+ }
+ memcpy(ctx->rec_seq, rec_seq, rec_seq_size);
+
+ sg_init_table(sw_ctx->sg_encrypted_data,
+ ARRAY_SIZE(sw_ctx->sg_encrypted_data));
+ sg_init_table(sw_ctx->sg_plaintext_data,
+ ARRAY_SIZE(sw_ctx->sg_plaintext_data));
+
+ sg_init_table(sw_ctx->sg_aead_in, 2);
+ sg_set_buf(&sw_ctx->sg_aead_in[0], sw_ctx->aad_space,
+ sizeof(sw_ctx->aad_space));
+ sg_unmark_end(&sw_ctx->sg_aead_in[1]);
+ sg_chain(sw_ctx->sg_aead_in, 2, sw_ctx->sg_plaintext_data);
+ sg_init_table(sw_ctx->sg_aead_out, 2);
+ sg_set_buf(&sw_ctx->sg_aead_out[0], sw_ctx->aad_space,
+ sizeof(sw_ctx->aad_space));
+ sg_unmark_end(&sw_ctx->sg_aead_out[1]);
+ sg_chain(sw_ctx->sg_aead_out, 2, sw_ctx->sg_encrypted_data);
+
+ if (!sw_ctx->aead_send) {
+ sw_ctx->aead_send = crypto_alloc_aead("gcm(aes)", 0, 0);
+ if (IS_ERR(sw_ctx->aead_send)) {
+ rc = PTR_ERR(sw_ctx->aead_send);
+ sw_ctx->aead_send = NULL;
+ goto free_rec_seq;
+ }
+ }
+
+ sk->sk_destruct = tls_sw_sk_destruct;
+ ctx->push_pending_record = tls_sw_push_pending_record;
+
+ memcpy(keyval, gcm_128_info->key, TLS_CIPHER_AES_GCM_128_KEY_SIZE);
+
+ rc = crypto_aead_setkey(sw_ctx->aead_send, keyval,
+ TLS_CIPHER_AES_GCM_128_KEY_SIZE);
+ if (rc)
+ goto free_aead;
+
+ rc = crypto_aead_setauthsize(sw_ctx->aead_send, ctx->tag_size);
+ if (!rc)
+ goto out;
+
+free_aead:
+ crypto_free_aead(sw_ctx->aead_send);
+ sw_ctx->aead_send = NULL;
+free_rec_seq:
+ kfree(ctx->rec_seq);
+ ctx->rec_seq = NULL;
+free_iv:
+ kfree(ctx->iv);
+ ctx->iv = NULL;
+out:
+ return rc;
+}
--
2.9.3
^ permalink raw reply related
* [PATCH v2 net-next 4/4] tls: Documentation
From: Dave Watson @ 2017-06-06 17:02 UTC (permalink / raw)
To: Ilya Lesokhin, Aviad Yehezkel, Boris Pismenny, Liran Liss,
Matan Barak, David Miller, netdev, Tom Herbert, herbert,
linux-crypto, Hannes Frederic Sowa, Eric Dumazet
Cc: Alexei Starovoitov, nmav, fridolin.pokorny
In-Reply-To: <cover.1496767702.git.davejwatson@fb.com>
Add documentation for the tcp ULP tls interface.
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Dave Watson <davejwatson@fb.com>
---
Documentation/networking/tls.txt | 135 +++++++++++++++++++++++++++++++++++++++
1 file changed, 135 insertions(+)
create mode 100644 Documentation/networking/tls.txt
diff --git a/Documentation/networking/tls.txt b/Documentation/networking/tls.txt
new file mode 100644
index 0000000..77ed006
--- /dev/null
+++ b/Documentation/networking/tls.txt
@@ -0,0 +1,135 @@
+Overview
+========
+
+Transport Layer Security (TLS) is a Upper Layer Protocol (ULP) that runs over
+TCP. TLS provides end-to-end data integrity and confidentiality.
+
+User interface
+==============
+
+Creating a TLS connection
+-------------------------
+
+First create a new TCP socket and set the TLS ULP.
+
+ sock = socket(AF_INET, SOCK_STREAM, 0);
+ setsockopt(sock, SOL_TCP, TCP_ULP, "tls", sizeof("tls"));
+
+Setting the TLS ULP allows us to set/get TLS socket options. Currently
+only the symmetric encryption is handled in the kernel. After the TLS
+handshake is complete, we have all the parameters required to move the
+data-path to the kernel. There is a separate socket option for moving
+the transmit and the receive into the kernel.
+
+ /* From linux/tls.h */
+ struct tls_crypto_info {
+ unsigned short version;
+ unsigned short cipher_type;
+ };
+
+ struct tls12_crypto_info_aes_gcm_128 {
+ struct tls_crypto_info info;
+ unsigned char iv[TLS_CIPHER_AES_GCM_128_IV_SIZE];
+ unsigned char key[TLS_CIPHER_AES_GCM_128_KEY_SIZE];
+ unsigned char salt[TLS_CIPHER_AES_GCM_128_SALT_SIZE];
+ unsigned char rec_seq[TLS_CIPHER_AES_GCM_128_REC_SEQ_SIZE];
+ };
+
+
+ struct tls12_crypto_info_aes_gcm_128 crypto_info;
+
+ crypto_info.info.version = TLS_1_2_VERSION;
+ crypto_info.info.cipher_type = TLS_CIPHER_AES_GCM_128;
+ memcpy(crypto_info.iv, iv_write, TLS_CIPHER_AES_GCM_128_IV_SIZE);
+ memcpy(crypto_info.rec_seq, seq_number_write,
+ TLS_CIPHER_AES_GCM_128_REC_SEQ_SIZE);
+ memcpy(crypto_info.key, cipher_key_write, TLS_CIPHER_AES_GCM_128_KEY_SIZE);
+ memcpy(crypto_info.salt, implicit_iv_write, TLS_CIPHER_AES_GCM_128_SALT_SIZE);
+
+ setsockopt(sock, SOL_TLS, TLS_TX, &crypto_info, sizeof(crypto_info));
+
+Sending TLS application data
+----------------------------
+
+After setting the TLS_TX socket option all application data sent over this
+socket is encrypted using TLS and the parameters provided in the socket option.
+For example, we can send an encrypted hello world record as follows:
+
+ const char *msg = "hello world\n";
+ send(sock, msg, strlen(msg));
+
+send() data is directly encrypted from the userspace buffer provided
+to the encrypted kernel send buffer if possible.
+
+The sendfile system call will send the file's data over TLS records of maximum
+length (2^14).
+
+ file = open(filename, O_RDONLY);
+ fstat(file, &stat);
+ sendfile(sock, file, &offset, stat.st_size);
+
+TLS records are created and sent after each send() call, unless
+MSG_MORE is passed. MSG_MORE will delay creation of a record until
+MSG_MORE is not passed, or the maximum record size is reached.
+
+The kernel will need to allocate a buffer for the encrypted data.
+This buffer is allocated at the time send() is called, such that
+either the entire send() call will return -ENOMEM (or block waiting
+for memory), or the encryption will always succeed. If send() returns
+-ENOMEM and some data was left on the socket buffer from a previous
+call using MSG_MORE, the MSG_MORE data is left on the socket buffer.
+
+Send TLS control messages
+-------------------------
+
+Other than application data, TLS has control messages such as alert
+messages (record type 21) and handshake messages (record type 22), etc.
+These messages can be sent over the socket by providing the TLS record type
+via a CMSG. For example the following function sends @data of @length bytes
+using a record of type @record_type.
+
+/* send TLS control message using record_type */
+ static int klts_send_ctrl_message(int sock, unsigned char record_type,
+ void *data, size_t length)
+ {
+ struct msghdr msg = {0};
+ int cmsg_len = sizeof(record_type);
+ struct cmsghdr *cmsg;
+ char buf[CMSG_SPACE(cmsg_len)];
+ struct iovec msg_iov; /* Vector of data to send/receive into. */
+
+ msg.msg_control = buf;
+ msg.msg_controllen = sizeof(buf);
+ cmsg = CMSG_FIRSTHDR(&msg);
+ cmsg->cmsg_level = SOL_TLS;
+ cmsg->cmsg_type = TLS_SET_RECORD_TYPE;
+ cmsg->cmsg_len = CMSG_LEN(cmsg_len);
+ *CMSG_DATA(cmsg) = record_type;
+ msg.msg_controllen = cmsg->cmsg_len;
+
+ msg_iov.iov_base = data;
+ msg_iov.iov_len = length;
+ msg.msg_iov = &msg_iov;
+ msg.msg_iovlen = 1;
+
+ return sendmsg(sock, &msg, 0);
+ }
+
+Control message data should be provided unencrypted, and will be
+encrypted by the kernel.
+
+Integrating in to userspace TLS library
+---------------------------------------
+
+At a high level, the kernel TLS ULP is a replacement for the record
+layer of a userspace TLS library.
+
+A patchset to OpenSSL to use ktls as the record layer is here:
+
+https://github.com/Mellanox/tls-openssl
+
+An example of calling send directly after a handshake using
+gnutls. Since it doesn't implement a full record layer, control
+messages are not supported:
+
+https://github.com/Mellanox/tls-af_ktls_tool
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 0/5] net: dsa: Multi-CPU ground work
From: Florian Fainelli @ 2017-06-06 17:01 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
Hi all,
This patch series prepares the ground for adding mutliple CPU port support to
DSA, and starts by removing redundant pieces of information such as
master_netdev which is cpu_dp->ethernet. Finally drivers are moved away from
directly accessing ds->dst->cpu_dp and use appropriate helper functions.
Note that if you have Device Tree blobs/platform configurations that are
currently listing multiple CPU ports, the proposed behavior in
dsa_ds_get_cpu_dp() will be to return the last bit set in ds->cpu_port_mask.
Future plans include:
- making dst->cpu_dp a flexible data structure (array, list, you name it)
- having the ability for drivers to return a default/preferred CPU port (if
necessary)
Florian Fainelli (5):
net: dsa: Remove master_netdev and use dst->cpu_dp->netdev
net: dsa: Relocate master ethtool operations
net: dsa: Associate slave network device with CPU port
net: dsa: Introduce dsa_dst_get_cpu_dp()
net: dsa: Stop accessing ds->dst->cpu_dp in drivers
drivers/net/dsa/b53/b53_common.c | 4 +--
drivers/net/dsa/bcm_sf2.c | 10 +++++---
drivers/net/dsa/mt7530.c | 6 +++--
drivers/net/dsa/mv88e6060.c | 3 ++-
drivers/net/dsa/qca8k.c | 2 +-
include/net/dsa.h | 29 +++++++++------------
net/dsa/dsa.c | 19 ++++----------
net/dsa/dsa2.c | 18 ++++++--------
net/dsa/dsa_priv.h | 10 ++++++++
net/dsa/legacy.c | 22 +++++++++-------
net/dsa/slave.c | 54 +++++++++++++++++++++-------------------
net/dsa/tag_brcm.c | 5 ++--
net/dsa/tag_ksz.c | 5 ++--
net/dsa/tag_qca.c | 3 ++-
net/dsa/tag_trailer.c | 5 ++--
15 files changed, 98 insertions(+), 97 deletions(-)
--
2.9.3
^ permalink raw reply
* [PATCH net-next 1/5] net: dsa: Remove master_netdev and use dst->cpu_dp->netdev
From: Florian Fainelli @ 2017-06-06 17:01 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
In preparation for supporting multiple CPU ports, remove
dst->master_netdev and ds->master_netdev and replace them with only one
instance of the common object we have for a port: struct
dsa_port::netdev. ds->master_netdev is currently write only and would be
helpful in the case where we have two switches, both with CPU ports, and
also connected within each other, which the multi-CPU port patch series
would address.
While at it, introduce a helper function used in net/dsa/slave.c to
immediately get a reference on the master network device called
dsa_master_netdev().
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/dsa/bcm_sf2.c | 4 ++--
drivers/net/dsa/mt7530.c | 4 ++--
include/net/dsa.h | 5 -----
net/dsa/dsa.c | 9 ++-------
net/dsa/dsa2.c | 18 +++++++-----------
net/dsa/dsa_priv.h | 5 +++++
net/dsa/legacy.c | 22 +++++++++++++---------
net/dsa/slave.c | 20 +++++++++-----------
8 files changed, 40 insertions(+), 47 deletions(-)
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 687a8bae5d73..76e98e8ed315 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -806,7 +806,7 @@ static int bcm_sf2_sw_resume(struct dsa_switch *ds)
static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = ds->dst[ds->index].master_netdev;
+ struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
struct ethtool_wolinfo pwol;
@@ -829,7 +829,7 @@ static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
static int bcm_sf2_sw_set_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = ds->dst[ds->index].master_netdev;
+ struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
s8 cpu_port = ds->dst->cpu_dp->index;
struct ethtool_wolinfo pwol;
diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 25e00d5e0eec..1e46418a3b74 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -912,11 +912,11 @@ mt7530_setup(struct dsa_switch *ds)
struct device_node *dn;
struct mt7530_dummy_poll p;
- /* The parent node of master_netdev which holds the common system
+ /* The parent node of cpu_dp->netdev which holds the common system
* controller also is the container for two GMACs nodes representing
* as two netdev instances.
*/
- dn = ds->master_netdev->dev.of_node->parent;
+ dn = ds->dst->cpu_dp->netdev->dev.of_node->parent;
priv->ethernet = syscon_node_to_regmap(dn);
if (IS_ERR(priv->ethernet))
return PTR_ERR(priv->ethernet);
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 2effb0af9d7c..b2fb53f5e28e 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -227,11 +227,6 @@ struct dsa_switch {
s8 rtable[DSA_MAX_SWITCHES];
/*
- * The lower device this switch uses to talk to the host
- */
- struct net_device *master_netdev;
-
- /*
* Slave mii_bus and devices for the individual ports.
*/
u32 dsa_port_mask;
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index fdc448b30e56..eaab1affeeeb 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -118,10 +118,7 @@ int dsa_cpu_port_ethtool_setup(struct dsa_port *cpu_dp)
struct net_device *master;
struct ethtool_ops *cpu_ops;
- master = ds->dst->master_netdev;
- if (ds->master_netdev)
- master = ds->master_netdev;
-
+ master = ds->dst->cpu_dp->netdev;
cpu_ops = devm_kzalloc(ds->dev, sizeof(*cpu_ops), GFP_KERNEL);
if (!cpu_ops)
return -ENOMEM;
@@ -142,9 +139,7 @@ void dsa_cpu_port_ethtool_restore(struct dsa_port *cpu_dp)
struct dsa_switch *ds = cpu_dp->ds;
struct net_device *master;
- master = ds->dst->master_netdev;
- if (ds->master_netdev)
- master = ds->master_netdev;
+ master = ds->dst->cpu_dp->netdev;
master->ethtool_ops = ds->dst->master_orig_ethtool_ops;
}
diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
index cd13bb54a30c..2674bdf03fef 100644
--- a/net/dsa/dsa2.c
+++ b/net/dsa/dsa2.c
@@ -337,7 +337,7 @@ static int dsa_ds_apply(struct dsa_switch_tree *dst, struct dsa_switch *ds)
return err;
if (ds->ops->set_addr) {
- err = ds->ops->set_addr(ds, dst->master_netdev->dev_addr);
+ err = ds->ops->set_addr(ds, dst->cpu_dp->netdev->dev_addr);
if (err < 0)
return err;
}
@@ -444,7 +444,7 @@ static int dsa_dst_apply(struct dsa_switch_tree *dst)
* sent to the tag format's receive function.
*/
wmb();
- dst->master_netdev->dsa_ptr = dst;
+ dst->cpu_dp->netdev->dsa_ptr = dst;
dst->applied = true;
return 0;
@@ -458,7 +458,7 @@ static void dsa_dst_unapply(struct dsa_switch_tree *dst)
if (!dst->applied)
return;
- dst->master_netdev->dsa_ptr = NULL;
+ dst->cpu_dp->netdev->dsa_ptr = NULL;
/* If we used a tagging format that doesn't have an ethertype
* field, make sure that all packets from this point get sent
@@ -502,14 +502,10 @@ static int dsa_cpu_parse(struct dsa_port *port, u32 index,
if (!ethernet_dev)
return -EPROBE_DEFER;
- if (!ds->master_netdev)
- ds->master_netdev = ethernet_dev;
-
- if (!dst->master_netdev)
- dst->master_netdev = ethernet_dev;
-
- if (!dst->cpu_dp)
+ if (!dst->cpu_dp) {
dst->cpu_dp = port;
+ dst->cpu_dp->netdev = ethernet_dev;
+ }
tag_protocol = ds->ops->get_tag_protocol(ds);
dst->tag_ops = dsa_resolve_tag_protocol(tag_protocol);
@@ -576,7 +572,7 @@ static int dsa_dst_parse(struct dsa_switch_tree *dst)
return err;
}
- if (!dst->master_netdev) {
+ if (!dst->cpu_dp->netdev) {
pr_warn("Tree has no master device\n");
return -EINVAL;
}
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 66ee248796c8..5c510f4ba0ce 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -183,4 +183,9 @@ extern const struct dsa_device_ops qca_netdev_ops;
/* tag_trailer.c */
extern const struct dsa_device_ops trailer_netdev_ops;
+static inline struct net_device *dsa_master_netdev(struct dsa_slave_priv *p)
+{
+ return p->dp->ds->dst->cpu_dp->netdev;
+}
+
#endif
diff --git a/net/dsa/legacy.c b/net/dsa/legacy.c
index d534d8f4b9cf..443ebf3b5e5b 100644
--- a/net/dsa/legacy.c
+++ b/net/dsa/legacy.c
@@ -101,9 +101,12 @@ static int dsa_switch_setup_one(struct dsa_switch *ds, struct device *parent)
struct dsa_switch_tree *dst = ds->dst;
struct dsa_chip_data *cd = ds->cd;
bool valid_name_found = false;
+ struct net_device *master;
int index = ds->index;
int i, ret;
+ master = dst->cpu_dp->netdev;
+
/*
* Validate supplied switch configuration.
*/
@@ -116,7 +119,7 @@ static int dsa_switch_setup_one(struct dsa_switch *ds, struct device *parent)
if (!strcmp(name, "cpu")) {
if (dst->cpu_dp) {
- netdev_err(dst->master_netdev,
+ netdev_err(master,
"multiple cpu ports?!\n");
return -EINVAL;
}
@@ -168,7 +171,7 @@ static int dsa_switch_setup_one(struct dsa_switch *ds, struct device *parent)
return ret;
if (ops->set_addr) {
- ret = ops->set_addr(ds, dst->master_netdev->dev_addr);
+ ret = ops->set_addr(ds, master->dev_addr);
if (ret < 0)
return ret;
}
@@ -195,14 +198,14 @@ static int dsa_switch_setup_one(struct dsa_switch *ds, struct device *parent)
ret = dsa_slave_create(ds, parent, i, cd->port_names[i]);
if (ret < 0)
- netdev_err(dst->master_netdev, "[%d]: can't create dsa slave device for port %d(%s): %d\n",
+ netdev_err(master, "[%d]: can't create dsa slave device for port %d(%s): %d\n",
index, i, cd->port_names[i], ret);
}
/* Perform configuration of the CPU and DSA ports */
ret = dsa_cpu_dsa_setups(ds, parent);
if (ret < 0)
- netdev_err(dst->master_netdev, "[%d] : can't configure CPU and DSA ports\n",
+ netdev_err(master, "[%d] : can't configure CPU and DSA ports\n",
index);
ret = dsa_cpu_port_ethtool_setup(ds->dst->cpu_dp);
@@ -217,6 +220,7 @@ dsa_switch_setup(struct dsa_switch_tree *dst, int index,
struct device *parent, struct device *host_dev)
{
struct dsa_chip_data *cd = dst->pd->chip + index;
+ struct net_device *master = dst->cpu_dp->netdev;
const struct dsa_switch_ops *ops;
struct dsa_switch *ds;
int ret;
@@ -228,11 +232,11 @@ dsa_switch_setup(struct dsa_switch_tree *dst, int index,
*/
ops = dsa_switch_probe(parent, host_dev, cd->sw_addr, &name, &priv);
if (!ops) {
- netdev_err(dst->master_netdev, "[%d]: could not detect attached switch\n",
+ netdev_err(master, "[%d]: could not detect attached switch\n",
index);
return ERR_PTR(-EINVAL);
}
- netdev_info(dst->master_netdev, "[%d]: detected a %s switch\n",
+ netdev_info(master, "[%d]: detected a %s switch\n",
index, name);
@@ -622,7 +626,7 @@ static int dsa_setup_dst(struct dsa_switch_tree *dst, struct net_device *dev,
unsigned configured = 0;
dst->pd = pd;
- dst->master_netdev = dev;
+ dst->cpu_dp->netdev = dev;
for (i = 0; i < pd->nr_chips; i++) {
struct dsa_switch *ds;
@@ -718,7 +722,7 @@ static void dsa_remove_dst(struct dsa_switch_tree *dst)
{
int i;
- dst->master_netdev->dsa_ptr = NULL;
+ dst->cpu_dp->netdev->dsa_ptr = NULL;
/* If we used a tagging format that doesn't have an ethertype
* field, make sure that all packets from this point get sent
@@ -735,7 +739,7 @@ static void dsa_remove_dst(struct dsa_switch_tree *dst)
dsa_cpu_port_ethtool_restore(dst->cpu_dp);
- dev_put(dst->master_netdev);
+ dev_put(dst->cpu_dp->netdev);
}
static int dsa_remove(struct platform_device *pdev)
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 1cfdb31a2f44..d52c9ceb0566 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -66,7 +66,7 @@ static int dsa_slave_get_iflink(const struct net_device *dev)
{
struct dsa_slave_priv *p = netdev_priv(dev);
- return p->dp->ds->dst->master_netdev->ifindex;
+ return dsa_master_netdev(p)->ifindex;
}
static int dsa_slave_open(struct net_device *dev)
@@ -74,7 +74,7 @@ static int dsa_slave_open(struct net_device *dev)
struct dsa_slave_priv *p = netdev_priv(dev);
struct dsa_port *dp = p->dp;
struct dsa_switch *ds = dp->ds;
- struct net_device *master = ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
u8 stp_state = dp->bridge_dev ? BR_STATE_BLOCKING : BR_STATE_FORWARDING;
int err;
@@ -127,7 +127,7 @@ static int dsa_slave_open(struct net_device *dev)
static int dsa_slave_close(struct net_device *dev)
{
struct dsa_slave_priv *p = netdev_priv(dev);
- struct net_device *master = p->dp->ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
struct dsa_switch *ds = p->dp->ds;
if (p->phy)
@@ -154,7 +154,7 @@ static int dsa_slave_close(struct net_device *dev)
static void dsa_slave_change_rx_flags(struct net_device *dev, int change)
{
struct dsa_slave_priv *p = netdev_priv(dev);
- struct net_device *master = p->dp->ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
if (change & IFF_ALLMULTI)
dev_set_allmulti(master, dev->flags & IFF_ALLMULTI ? 1 : -1);
@@ -165,7 +165,7 @@ static void dsa_slave_change_rx_flags(struct net_device *dev, int change)
static void dsa_slave_set_rx_mode(struct net_device *dev)
{
struct dsa_slave_priv *p = netdev_priv(dev);
- struct net_device *master = p->dp->ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
dev_mc_sync(master, dev);
dev_uc_sync(master, dev);
@@ -174,7 +174,7 @@ static void dsa_slave_set_rx_mode(struct net_device *dev)
static int dsa_slave_set_mac_address(struct net_device *dev, void *a)
{
struct dsa_slave_priv *p = netdev_priv(dev);
- struct net_device *master = p->dp->ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
struct sockaddr *addr = a;
int err;
@@ -375,7 +375,7 @@ static netdev_tx_t dsa_slave_xmit(struct sk_buff *skb, struct net_device *dev)
/* Queue the SKB for transmission on the parent interface, but
* do not modify its EtherType
*/
- nskb->dev = p->dp->ds->dst->master_netdev;
+ nskb->dev = dsa_master_netdev(p);
dev_queue_xmit(nskb);
return NETDEV_TX_OK;
@@ -684,7 +684,7 @@ static int dsa_slave_netpoll_setup(struct net_device *dev,
{
struct dsa_slave_priv *p = netdev_priv(dev);
struct dsa_switch *ds = p->dp->ds;
- struct net_device *master = ds->dst->master_netdev;
+ struct net_device *master = dsa_master_netdev(p);
struct netpoll *netpoll;
int err = 0;
@@ -1141,9 +1141,7 @@ int dsa_slave_create(struct dsa_switch *ds, struct device *parent,
struct dsa_slave_priv *p;
int ret;
- master = ds->dst->master_netdev;
- if (ds->master_netdev)
- master = ds->master_netdev;
+ master = ds->dst->cpu_dp->netdev;
slave_dev = alloc_netdev(sizeof(struct dsa_slave_priv), name,
NET_NAME_UNKNOWN, ether_setup);
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 4/5] net: dsa: Introduce dsa_dst_get_cpu_dp()
From: Florian Fainelli @ 2017-06-06 17:02 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
Introduce a helper function which will return a reference to the CPU
port used in a dsa_switch_tree. Right now this is a singleton, but this
will change once we introduce multi-CPU port support, so ease the
transition by converting the affected code paths.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
net/dsa/dsa_priv.h | 5 +++++
net/dsa/slave.c | 31 ++++++++++++++++---------------
net/dsa/tag_brcm.c | 5 ++---
net/dsa/tag_ksz.c | 5 ++---
net/dsa/tag_qca.c | 3 ++-
net/dsa/tag_trailer.c | 5 ++---
6 files changed, 29 insertions(+), 25 deletions(-)
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 7c2326f3b538..49b4b047aed0 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -188,4 +188,9 @@ static inline struct net_device *dsa_master_netdev(struct dsa_slave_priv *p)
return p->dp->cpu_dp->netdev;
}
+static inline struct dsa_port *dsa_dst_get_cpu_dp(struct dsa_switch_tree *dst)
+{
+ return dst->cpu_dp;
+}
+
#endif
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index de1ab41cfd38..a73c1de398b5 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -519,14 +519,14 @@ static void dsa_cpu_port_get_ethtool_stats(struct net_device *dev,
uint64_t *data)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds = dst->cpu_dp->ds;
- s8 cpu_port = dst->cpu_dp->index;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
+ s8 cpu_port = cpu_dp->index;
int count = 0;
- if (dst->cpu_dp->ethtool_ops.get_sset_count) {
- count = dst->cpu_dp->ethtool_ops.get_sset_count(dev,
- ETH_SS_STATS);
- dst->cpu_dp->ethtool_ops.get_ethtool_stats(dev, stats, data);
+ if (cpu_dp->ethtool_ops.get_sset_count) {
+ count = cpu_dp->ethtool_ops.get_sset_count(dev, ETH_SS_STATS);
+ cpu_dp->ethtool_ops.get_ethtool_stats(dev, stats, data);
}
if (ds->ops->get_ethtool_stats)
@@ -536,11 +536,12 @@ static void dsa_cpu_port_get_ethtool_stats(struct net_device *dev,
static int dsa_cpu_port_get_sset_count(struct net_device *dev, int sset)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds = dst->cpu_dp->ds;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
int count = 0;
- if (dst->cpu_dp->ethtool_ops.get_sset_count)
- count += dst->cpu_dp->ethtool_ops.get_sset_count(dev, sset);
+ if (cpu_dp->ethtool_ops.get_sset_count)
+ count += cpu_dp->ethtool_ops.get_sset_count(dev, sset);
if (sset == ETH_SS_STATS && ds->ops->get_sset_count)
count += ds->ops->get_sset_count(ds);
@@ -552,8 +553,9 @@ static void dsa_cpu_port_get_strings(struct net_device *dev,
uint32_t stringset, uint8_t *data)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds = dst->cpu_dp->ds;
- s8 cpu_port = dst->cpu_dp->index;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
+ s8 cpu_port = cpu_dp->index;
int len = ETH_GSTRING_LEN;
int mcount = 0, count;
unsigned int i;
@@ -564,10 +566,9 @@ static void dsa_cpu_port_get_strings(struct net_device *dev,
/* We do not want to be NULL-terminated, since this is a prefix */
pfx[sizeof(pfx) - 1] = '_';
- if (dst->cpu_dp->ethtool_ops.get_sset_count) {
- mcount = dst->cpu_dp->ethtool_ops.get_sset_count(dev,
- ETH_SS_STATS);
- dst->cpu_dp->ethtool_ops.get_strings(dev, stringset, data);
+ if (cpu_dp->ethtool_ops.get_sset_count) {
+ mcount = cpu_dp->ethtool_ops.get_sset_count(dev, ETH_SS_STATS);
+ cpu_dp->ethtool_ops.get_strings(dev, stringset, data);
}
if (stringset == ETH_SS_STATS && ds->ops->get_strings) {
diff --git a/net/dsa/tag_brcm.c b/net/dsa/tag_brcm.c
index c03860907f28..d7ef2b35e61e 100644
--- a/net/dsa/tag_brcm.c
+++ b/net/dsa/tag_brcm.c
@@ -93,12 +93,11 @@ static struct sk_buff *brcm_tag_rcv(struct sk_buff *skb, struct net_device *dev,
struct net_device *orig_dev)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
int source_port;
u8 *brcm_tag;
- ds = dst->cpu_dp->ds;
-
if (unlikely(!pskb_may_pull(skb, BRCM_TAG_LEN)))
return NULL;
diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index b94a334a1d02..c41a24e83e83 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -75,12 +75,11 @@ static struct sk_buff *ksz_rcv(struct sk_buff *skb, struct net_device *dev,
struct net_device *orig_dev)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
u8 *tag;
int source_port;
- ds = dst->cpu_dp->ds;
-
tag = skb_tail_pointer(skb) - KSZ_EGRESS_TAG_LEN;
source_port = tag[0] & 7;
diff --git a/net/dsa/tag_qca.c b/net/dsa/tag_qca.c
index 4f43cf0b4eff..73244ae388df 100644
--- a/net/dsa/tag_qca.c
+++ b/net/dsa/tag_qca.c
@@ -67,6 +67,7 @@ static struct sk_buff *qca_tag_rcv(struct sk_buff *skb, struct net_device *dev,
struct net_device *orig_dev)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
struct dsa_switch *ds;
u8 ver;
int port;
@@ -95,7 +96,7 @@ static struct sk_buff *qca_tag_rcv(struct sk_buff *skb, struct net_device *dev,
/* This protocol doesn't support cascading multiple switches so it's
* safe to assume the switch is first in the tree
*/
- ds = dst->cpu_dp->ds;
+ ds = cpu_dp->ds;
if (!ds)
return NULL;
diff --git a/net/dsa/tag_trailer.c b/net/dsa/tag_trailer.c
index b4f6db094409..68f664cb7517 100644
--- a/net/dsa/tag_trailer.c
+++ b/net/dsa/tag_trailer.c
@@ -61,12 +61,11 @@ static struct sk_buff *trailer_rcv(struct sk_buff *skb, struct net_device *dev,
struct net_device *orig_dev)
{
struct dsa_switch_tree *dst = dev->dsa_ptr;
- struct dsa_switch *ds;
+ struct dsa_port *cpu_dp = dsa_dst_get_cpu_dp(dst);
+ struct dsa_switch *ds = cpu_dp->ds;
u8 *trailer;
int source_port;
- ds = dst->cpu_dp->ds;
-
if (skb_linearize(skb))
return NULL;
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 5/5] net: dsa: Stop accessing ds->dst->cpu_dp in drivers
From: Florian Fainelli @ 2017-06-06 17:02 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
Out of the few drivers that do access ds->dst->cpu_dp, there is only a
handful for which we cannot substitute that for either an existing and
equivalent piece of information (b53, bcm_sf2, qca8k), and there is only
one for which we need to introduce a helper: mt7530. We do introduce
dsa_ds_get_cpu_dp() which reads the CPU port from ds->cpu_port_mask.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/dsa/b53/b53_common.c | 4 ++--
drivers/net/dsa/bcm_sf2.c | 10 ++++++----
drivers/net/dsa/mt7530.c | 4 +++-
drivers/net/dsa/mv88e6060.c | 3 ++-
drivers/net/dsa/qca8k.c | 2 +-
include/net/dsa.h | 6 ++++++
6 files changed, 20 insertions(+), 9 deletions(-)
diff --git a/drivers/net/dsa/b53/b53_common.c b/drivers/net/dsa/b53/b53_common.c
index e68d368e20ac..faec6fcacd31 100644
--- a/drivers/net/dsa/b53/b53_common.c
+++ b/drivers/net/dsa/b53/b53_common.c
@@ -1341,7 +1341,7 @@ EXPORT_SYMBOL(b53_fdb_dump);
int b53_br_join(struct dsa_switch *ds, int port, struct net_device *br)
{
struct b53_device *dev = ds->priv;
- s8 cpu_port = ds->dst->cpu_dp->index;
+ s8 cpu_port = dev->cpu_port;
u16 pvlan, reg;
unsigned int i;
@@ -1387,7 +1387,7 @@ void b53_br_leave(struct dsa_switch *ds, int port, struct net_device *br)
{
struct b53_device *dev = ds->priv;
struct b53_vlan *vl = &dev->vlans[0];
- s8 cpu_port = ds->dst->cpu_dp->index;
+ s8 cpu_port = dev->cpu_port;
unsigned int i;
u16 pvlan, reg, pvid;
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 76e98e8ed315..9744100d0276 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -227,7 +227,7 @@ static int bcm_sf2_port_setup(struct dsa_switch *ds, int port,
struct phy_device *phy)
{
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
- s8 cpu_port = ds->dst->cpu_dp->index;
+ s8 cpu_port = priv->dev->cpu_port;
unsigned int i;
u32 reg;
@@ -806,8 +806,9 @@ static int bcm_sf2_sw_resume(struct dsa_switch *ds)
static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
+ struct dsa_port *cpu_dp = ds->ports[port].cpu_dp;
+ struct net_device *p = cpu_dp->netdev;
struct ethtool_wolinfo pwol;
/* Get the parent device WoL settings */
@@ -829,9 +830,10 @@ static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
static int bcm_sf2_sw_set_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
- s8 cpu_port = ds->dst->cpu_dp->index;
+ struct dsa_port *cpu_dp = ds->ports[port].cpu_dp;
+ struct net_device *p = cpu_dp->netdev;
+ s8 cpu_port = cpu_dp->index;
struct ethtool_wolinfo pwol;
p->ethtool_ops->get_wol(p, &pwol);
diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 1e46418a3b74..9b1b76c7b927 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -907,6 +907,7 @@ static int
mt7530_setup(struct dsa_switch *ds)
{
struct mt7530_priv *priv = ds->priv;
+ struct dsa_port *cpu_dp;
int ret, i;
u32 id, val;
struct device_node *dn;
@@ -916,7 +917,8 @@ mt7530_setup(struct dsa_switch *ds)
* controller also is the container for two GMACs nodes representing
* as two netdev instances.
*/
- dn = ds->dst->cpu_dp->netdev->dev.of_node->parent;
+ cpu_dp = dsa_ds_get_cpu_dp(ds);
+ dn = cpu_dp->netdev->dev.of_node->parent;
priv->ethernet = syscon_node_to_regmap(dn);
if (IS_ERR(priv->ethernet))
return PTR_ERR(priv->ethernet);
diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c
index dce7fa57eb55..76fcd4350679 100644
--- a/drivers/net/dsa/mv88e6060.c
+++ b/drivers/net/dsa/mv88e6060.c
@@ -153,6 +153,7 @@ static int mv88e6060_setup_global(struct dsa_switch *ds)
static int mv88e6060_setup_port(struct dsa_switch *ds, int p)
{
+ struct dsa_port *cpu_dp = dsa_ds_get_cpu_dp(ds);
int addr = REG_PORT(p);
/* Do not force flow control, disable Ingress and Egress
@@ -176,7 +177,7 @@ static int mv88e6060_setup_port(struct dsa_switch *ds, int p)
((p & 0xf) << PORT_VLAN_MAP_DBNUM_SHIFT) |
(dsa_is_cpu_port(ds, p) ?
ds->enabled_port_mask :
- BIT(ds->dst->cpu_dp->index)));
+ BIT(cpu_dp->index)));
/* Port Association Vector: when learning source addresses
* of packets, add the address to the address database using
diff --git a/drivers/net/dsa/qca8k.c b/drivers/net/dsa/qca8k.c
index b3bee7eab45f..68b45298f6d1 100644
--- a/drivers/net/dsa/qca8k.c
+++ b/drivers/net/dsa/qca8k.c
@@ -506,7 +506,7 @@ qca8k_setup(struct dsa_switch *ds)
pr_warn("regmap initialization failed");
/* Initialize CPU port pad mode (xMII type, delays...) */
- phy_mode = of_get_phy_mode(ds->dst->cpu_dp->dn);
+ phy_mode = of_get_phy_mode(ds->ports[QCA8K_CPU_PORT].dn);
if (phy_mode < 0) {
pr_err("Can't find phy-mode for master device\n");
return phy_mode;
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 58969b9a090c..264cb3f93764 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -11,6 +11,7 @@
#ifndef __LINUX_NET_DSA_H
#define __LINUX_NET_DSA_H
+#include <linux/bitops.h>
#include <linux/if.h>
#include <linux/if_ether.h>
#include <linux/list.h>
@@ -261,6 +262,11 @@ static inline bool dsa_is_port_initialized(struct dsa_switch *ds, int p)
return ds->enabled_port_mask & (1 << p) && ds->ports[p].netdev;
}
+static inline struct dsa_port *dsa_ds_get_cpu_dp(struct dsa_switch *ds)
+{
+ return &ds->ports[fls(ds->cpu_port_mask) - 1];
+}
+
static inline u8 dsa_upstream_port(struct dsa_switch *ds)
{
struct dsa_switch_tree *dst = ds->dst;
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 3/5] net: dsa: Associate slave network device with CPU port
From: Florian Fainelli @ 2017-06-06 17:02 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
In preparation for supporting multiple CPU ports with DSA, have the
dsa_slave_priv structure know which CPU it is associated with. This will
be important in order to make sure the correct CPU is used for
transmission of the frames. If not for functional reasons, for
performance (e.g: load balancing) and forwarding decisions.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
include/net/dsa.h | 1 +
net/dsa/dsa_priv.h | 2 +-
net/dsa/slave.c | 5 ++++-
3 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 7e93869819f9..58969b9a090c 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -171,6 +171,7 @@ struct dsa_port {
struct dsa_switch *ds;
unsigned int index;
const char *name;
+ struct dsa_port *cpu_dp;
struct net_device *netdev;
struct device_node *dn;
unsigned int ageing_time;
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 5c510f4ba0ce..7c2326f3b538 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -185,7 +185,7 @@ extern const struct dsa_device_ops trailer_netdev_ops;
static inline struct net_device *dsa_master_netdev(struct dsa_slave_priv *p)
{
- return p->dp->ds->dst->cpu_dp->netdev;
+ return p->dp->cpu_dp->netdev;
}
#endif
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index ea4ed0285922..de1ab41cfd38 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -1139,9 +1139,11 @@ int dsa_slave_create(struct dsa_switch *ds, struct device *parent,
struct net_device *master;
struct net_device *slave_dev;
struct dsa_slave_priv *p;
+ struct dsa_port *cpu_dp;
int ret;
- master = ds->dst->cpu_dp->netdev;
+ cpu_dp = ds->dst->cpu_dp;
+ master = cpu_dp->netdev;
slave_dev = alloc_netdev(sizeof(struct dsa_slave_priv), name,
NET_NAME_UNKNOWN, ether_setup);
@@ -1176,6 +1178,7 @@ int dsa_slave_create(struct dsa_switch *ds, struct device *parent,
p->old_duplex = -1;
ds->ports[port].netdev = slave_dev;
+ p->dp->cpu_dp = cpu_dp;
ret = register_netdev(slave_dev);
if (ret) {
netdev_err(master, "error %d registering interface %s\n",
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 2/5] net: dsa: Relocate master ethtool operations
From: Florian Fainelli @ 2017-06-06 17:01 UTC (permalink / raw)
To: netdev; +Cc: andrew, vivien.didelot, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
Relocate master_ethtool_ops and master_orig_ethtool_ops into struct
dsa_port in order to be both consistent, and make things self contained
within the dsa_port structure.
This is a preliminary change to supporting multiple CPU port interfaces.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
include/net/dsa.h | 17 +++++------------
net/dsa/dsa.c | 16 ++++++----------
net/dsa/slave.c | 16 ++++++++--------
3 files changed, 19 insertions(+), 30 deletions(-)
diff --git a/include/net/dsa.h b/include/net/dsa.h
index b2fb53f5e28e..7e93869819f9 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -122,12 +122,6 @@ struct dsa_switch_tree {
*/
struct dsa_platform_data *pd;
- /*
- * Reference to network device to use, and which tagging
- * protocol to use.
- */
- struct net_device *master_netdev;
-
/* Copy of tag_ops->rcv for faster access in hot path */
struct sk_buff * (*rcv)(struct sk_buff *skb,
struct net_device *dev,
@@ -135,12 +129,6 @@ struct dsa_switch_tree {
struct net_device *orig_dev);
/*
- * Original copy of the master netdev ethtool_ops
- */
- struct ethtool_ops master_ethtool_ops;
- const struct ethtool_ops *master_orig_ethtool_ops;
-
- /*
* The switch port to which the CPU is attached.
*/
struct dsa_port *cpu_dp;
@@ -189,6 +177,11 @@ struct dsa_port {
u8 stp_state;
struct net_device *bridge_dev;
struct devlink_port devlink_port;
+ /*
+ * Original copy of the master netdev ethtool_ops
+ */
+ struct ethtool_ops ethtool_ops;
+ const struct ethtool_ops *orig_ethtool_ops;
};
struct dsa_switch {
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index eaab1affeeeb..2665a66e833d 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -118,15 +118,16 @@ int dsa_cpu_port_ethtool_setup(struct dsa_port *cpu_dp)
struct net_device *master;
struct ethtool_ops *cpu_ops;
- master = ds->dst->cpu_dp->netdev;
+ master = cpu_dp->netdev;
+
cpu_ops = devm_kzalloc(ds->dev, sizeof(*cpu_ops), GFP_KERNEL);
if (!cpu_ops)
return -ENOMEM;
- memcpy(&ds->dst->master_ethtool_ops, master->ethtool_ops,
+ memcpy(&cpu_dp->ethtool_ops, master->ethtool_ops,
sizeof(struct ethtool_ops));
- ds->dst->master_orig_ethtool_ops = master->ethtool_ops;
- memcpy(cpu_ops, &ds->dst->master_ethtool_ops,
+ cpu_dp->orig_ethtool_ops = master->ethtool_ops;
+ memcpy(cpu_ops, &cpu_dp->ethtool_ops,
sizeof(struct ethtool_ops));
dsa_cpu_port_ethtool_init(cpu_ops);
master->ethtool_ops = cpu_ops;
@@ -136,12 +137,7 @@ int dsa_cpu_port_ethtool_setup(struct dsa_port *cpu_dp)
void dsa_cpu_port_ethtool_restore(struct dsa_port *cpu_dp)
{
- struct dsa_switch *ds = cpu_dp->ds;
- struct net_device *master;
-
- master = ds->dst->cpu_dp->netdev;
-
- master->ethtool_ops = ds->dst->master_orig_ethtool_ops;
+ cpu_dp->netdev->ethtool_ops = cpu_dp->orig_ethtool_ops;
}
void dsa_cpu_dsa_destroy(struct dsa_port *port)
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index d52c9ceb0566..ea4ed0285922 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -523,10 +523,10 @@ static void dsa_cpu_port_get_ethtool_stats(struct net_device *dev,
s8 cpu_port = dst->cpu_dp->index;
int count = 0;
- if (dst->master_ethtool_ops.get_sset_count) {
- count = dst->master_ethtool_ops.get_sset_count(dev,
+ if (dst->cpu_dp->ethtool_ops.get_sset_count) {
+ count = dst->cpu_dp->ethtool_ops.get_sset_count(dev,
ETH_SS_STATS);
- dst->master_ethtool_ops.get_ethtool_stats(dev, stats, data);
+ dst->cpu_dp->ethtool_ops.get_ethtool_stats(dev, stats, data);
}
if (ds->ops->get_ethtool_stats)
@@ -539,8 +539,8 @@ static int dsa_cpu_port_get_sset_count(struct net_device *dev, int sset)
struct dsa_switch *ds = dst->cpu_dp->ds;
int count = 0;
- if (dst->master_ethtool_ops.get_sset_count)
- count += dst->master_ethtool_ops.get_sset_count(dev, sset);
+ if (dst->cpu_dp->ethtool_ops.get_sset_count)
+ count += dst->cpu_dp->ethtool_ops.get_sset_count(dev, sset);
if (sset == ETH_SS_STATS && ds->ops->get_sset_count)
count += ds->ops->get_sset_count(ds);
@@ -564,10 +564,10 @@ static void dsa_cpu_port_get_strings(struct net_device *dev,
/* We do not want to be NULL-terminated, since this is a prefix */
pfx[sizeof(pfx) - 1] = '_';
- if (dst->master_ethtool_ops.get_sset_count) {
- mcount = dst->master_ethtool_ops.get_sset_count(dev,
+ if (dst->cpu_dp->ethtool_ops.get_sset_count) {
+ mcount = dst->cpu_dp->ethtool_ops.get_sset_count(dev,
ETH_SS_STATS);
- dst->master_ethtool_ops.get_strings(dev, stringset, data);
+ dst->cpu_dp->ethtool_ops.get_strings(dev, stringset, data);
}
if (stringset == ETH_SS_STATS && ds->ops->get_strings) {
--
2.9.3
^ permalink raw reply related
* Re: [PATCH net-next 0/9] s390/net updates
From: David Miller @ 2017-06-06 17:06 UTC (permalink / raw)
To: jwi; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, raspl, ubraun
In-Reply-To: <20170606123350.13781-1-jwi@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Date: Tue, 6 Jun 2017 14:33:41 +0200
> please apply the following qeth updates for net-next.
>
> Aside from some janitorial changes, this adds early setup for virtualized
> HiperSockets devices - building upon the code that landed via -net earlier.
Series applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next 1/5] net: dsa: Remove master_netdev and use dst->cpu_dp->netdev
From: Vivien Didelot @ 2017-06-06 17:24 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-2-f.fainelli@gmail.com>
Hi Florian,
Florian Fainelli <f.fainelli@gmail.com> writes:
> - struct net_device *p = ds->dst[ds->index].master_netdev;
> + struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
ds->dst->cpu_dp->netdev
ds->dst is not an array anymore, lucky sf2 switch index is always 0 ;-)
> struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
> struct ethtool_wolinfo pwol;
>
> @@ -829,7 +829,7 @@ static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
> static int bcm_sf2_sw_set_wol(struct dsa_switch *ds, int port,
> struct ethtool_wolinfo *wol)
> {
> - struct net_device *p = ds->dst[ds->index].master_netdev;
> + struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
same here.
Thanks,
Vivien
^ permalink raw reply
* Re: [PATCH net-next 2/5] net: dsa: Relocate master ethtool operations
From: Vivien Didelot @ 2017-06-06 17:28 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-3-f.fainelli@gmail.com>
Florian Fainelli <f.fainelli@gmail.com> writes:
> Relocate master_ethtool_ops and master_orig_ethtool_ops into struct
> dsa_port in order to be both consistent, and make things self contained
> within the dsa_port structure.
>
> This is a preliminary change to supporting multiple CPU port interfaces.
>
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
^ permalink raw reply
* Re: [PATCH net-next 3/5] net: dsa: Associate slave network device with CPU port
From: Vivien Didelot @ 2017-06-06 17:37 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-4-f.fainelli@gmail.com>
Florian Fainelli <f.fainelli@gmail.com> writes:
> In preparation for supporting multiple CPU ports with DSA, have the
> dsa_slave_priv structure know which CPU it is associated with. This will
dsa_port
> be important in order to make sure the correct CPU is used for
> transmission of the frames. If not for functional reasons, for
> performance (e.g: load balancing) and forwarding decisions.
>
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
> include/net/dsa.h | 1 +
> net/dsa/dsa_priv.h | 2 +-
> net/dsa/slave.c | 5 ++++-
> 3 files changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/include/net/dsa.h b/include/net/dsa.h
> index 7e93869819f9..58969b9a090c 100644
> --- a/include/net/dsa.h
> +++ b/include/net/dsa.h
> @@ -171,6 +171,7 @@ struct dsa_port {
> struct dsa_switch *ds;
> unsigned int index;
> const char *name;
> + struct dsa_port *cpu_dp;
> struct net_device *netdev;
> struct device_node *dn;
> unsigned int ageing_time;
> diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
> index 5c510f4ba0ce..7c2326f3b538 100644
> --- a/net/dsa/dsa_priv.h
> +++ b/net/dsa/dsa_priv.h
> @@ -185,7 +185,7 @@ extern const struct dsa_device_ops trailer_netdev_ops;
>
> static inline struct net_device *dsa_master_netdev(struct dsa_slave_priv *p)
> {
> - return p->dp->ds->dst->cpu_dp->netdev;
> + return p->dp->cpu_dp->netdev;
> }
>
> #endif
> diff --git a/net/dsa/slave.c b/net/dsa/slave.c
> index ea4ed0285922..de1ab41cfd38 100644
> --- a/net/dsa/slave.c
> +++ b/net/dsa/slave.c
> @@ -1139,9 +1139,11 @@ int dsa_slave_create(struct dsa_switch *ds, struct device *parent,
> struct net_device *master;
> struct net_device *slave_dev;
> struct dsa_slave_priv *p;
> + struct dsa_port *cpu_dp;
> int ret;
>
> - master = ds->dst->cpu_dp->netdev;
> + cpu_dp = ds->dst->cpu_dp;
> + master = cpu_dp->netdev;
You may assign them when declaring them, but no big deal.
>
> slave_dev = alloc_netdev(sizeof(struct dsa_slave_priv), name,
> NET_NAME_UNKNOWN, ether_setup);
> @@ -1176,6 +1178,7 @@ int dsa_slave_create(struct dsa_switch *ds, struct device *parent,
> p->old_duplex = -1;
>
> ds->ports[port].netdev = slave_dev;
> + p->dp->cpu_dp = cpu_dp;
> ret = register_netdev(slave_dev);
> if (ret) {
> netdev_err(master, "error %d registering interface %s\n",
Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
^ permalink raw reply
* Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
From: Alexei Starovoitov @ 2017-06-06 17:42 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: Ilan Tayari, Saeed Mahameed, David S. Miller, Doug Ledford,
netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
jsorensen-b10kYP2dOMg@public.gmane.org, Andy Shevchenko,
linux-fpga-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Alan Tull,
yi1.li-VuQAYsv1563Yd54FQh9/CA@public.gmane.org, Boris Pismenny
In-Reply-To: <20170606161709.GA8671-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
On Tue, Jun 06, 2017 at 10:17:09AM -0600, Jason Gunthorpe wrote:
> On Tue, Jun 06, 2017 at 06:52:15AM +0000, Ilan Tayari wrote:
>
> > So neither the host stack nor the network are aware of them.
> > They exist momentarily only on the internal traces on the board and not
> > anywhere else.
>
> Is that really true? If you are creating rocee QPs' then the RDMA
> stack sees this stuff and now we have buried a RDMA ULP inside an
> ethernet driver which seems really wonky..
>
> > I don't mind explaining further, but I think you will just see it in the
> > patchset when we submit.
>
> You described exactly what I thought.. I just disagree with you that
> an ethernet connected and controlled IP accelerator is 'part of the
> NIC', even if it happens to be colocated on the same circuit board.
+1
what Ilan described is a kernel bypass done by hw.
This is non starter in production. Same as eswitch this fpga is not
represented as a kernel object, there is no way to debug it.
NIC crafts roce packets back and forth?!
so it's like rdma, but without using kernel rdma stack?
When hw ipsec or tls will mysteriously drop or mangle the packets
how this can be debugged? Does fpga have attached ddr to
store/forward the packets? How memory issues will be reported?
No MCE errors ever? Buffer overflow? How many receive queues inside fpga?
How health check of fgpa itself will be done? Through roce packets?
I would buy the lack of kernel visibility if this fpga+nic combo
was a prototype, but it's being presented as a production device
with subsequent changes to core networking stack and that's where
I have a problem with its sw architecture.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
From: David Miller @ 2017-06-06 17:47 UTC (permalink / raw)
To: alexei.starovoitov-Re5JQEeQqe8AvxtiuMwx3w
Cc: jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/,
ilant-VPRAkNaXOzVWk0Htik3J/w,
saeedm-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb,
dledford-H+wXaHxf7aLQT0dZR+AlfA, netdev-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA, jsorensen-b10kYP2dOMg,
andy.shevchenko-Re5JQEeQqe8AvxtiuMwx3w,
linux-fpga-u79uwXL29TY76Z2rM5mHXA,
atull-yzvPICuk2ABMcg4IHK0kFoH6Mc4MB0Vx,
yi1.li-VuQAYsv1563Yd54FQh9/CA, borisp-VPRAkNaXOzVWk0Htik3J/w
In-Reply-To: <20170606174233.w377ctwtapzccsk7-+o4/htvd0TCa6kscz5V53/3mLCh9rsb+VpNB7YpNyf8@public.gmane.org>
From: Alexei Starovoitov <alexei.starovoitov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date: Tue, 6 Jun 2017 10:42:35 -0700
> so it's like rdma, but without using kernel rdma stack?
No sockets here, just transformation rules. It's like offloading
a complex TC rule to hardware version of that transformation.
Yes, there is state, but I argue that it is no different than TC
offloading rules. What if TC had "hash" and "crypt" operations
and we attached them to appropriate u32 matches? You wouldn't
be able to tell the difference.
I think you are way over-obsessed with this FPGA offload thing,
quite frankly.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH net-next 4/5] net: dsa: Introduce dsa_dst_get_cpu_dp()
From: Vivien Didelot @ 2017-06-06 17:52 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-5-f.fainelli@gmail.com>
Florian Fainelli <f.fainelli@gmail.com> writes:
> Introduce a helper function which will return a reference to the CPU
> port used in a dsa_switch_tree. Right now this is a singleton, but this
> will change once we introduce multi-CPU port support, so ease the
> transition by converting the affected code paths.
>
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
^ permalink raw reply
* Re: [PATCH net-next 1/5] net: dsa: Remove master_netdev and use dst->cpu_dp->netdev
From: Florian Fainelli @ 2017-06-06 18:09 UTC (permalink / raw)
To: Vivien Didelot, netdev; +Cc: andrew, john, davem
In-Reply-To: <87d1ahjagr.fsf@weeman.i-did-not-set--mail-host-address--so-tickle-me>
On 06/06/2017 10:24 AM, Vivien Didelot wrote:
> Hi Florian,
>
> Florian Fainelli <f.fainelli@gmail.com> writes:
>
>> - struct net_device *p = ds->dst[ds->index].master_netdev;
>> + struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
>
> ds->dst->cpu_dp->netdev
>
> ds->dst is not an array anymore, lucky sf2 switch index is always 0 ;-)
>
>> struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
>> struct ethtool_wolinfo pwol;
>>
>> @@ -829,7 +829,7 @@ static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
>> static int bcm_sf2_sw_set_wol(struct dsa_switch *ds, int port,
>> struct ethtool_wolinfo *wol)
>> {
>> - struct net_device *p = ds->dst[ds->index].master_netdev;
>> + struct net_device *p = ds->dst[ds->index].cpu_dp->netdev;
>
> same here.
It's changed in patch 5, so I did not bother doing an intermediate
change considering we ditch this eventually.
Thanks!
>
> Thanks,
>
> Vivien
>
--
Florian
^ permalink raw reply
* Re: [PATCH net-next 5/5] net: dsa: Stop accessing ds->dst->cpu_dp in drivers
From: Vivien Didelot @ 2017-06-06 18:09 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-6-f.fainelli@gmail.com>
Florian Fainelli <f.fainelli@gmail.com> writes:
> - phy_mode = of_get_phy_mode(ds->dst->cpu_dp->dn);
> + phy_mode = of_get_phy_mode(ds->ports[QCA8K_CPU_PORT].dn);
Is it necessary to use QCA8K_CPU_PORT?
> +static inline struct dsa_port *dsa_ds_get_cpu_dp(struct dsa_switch *ds)
> +{
> + return &ds->ports[fls(ds->cpu_port_mask) - 1];
> +}
Wouldn't it be better to return the CPU port for a given port?
Something like return ds->ports[port].cpu_dp, so that we ease the
introduction of multiple CPU port a bit more?
Thanks,
Vivien
^ permalink raw reply
* Re: [PATCH net-next 5/5] net: dsa: Stop accessing ds->dst->cpu_dp in drivers
From: Florian Fainelli @ 2017-06-06 18:12 UTC (permalink / raw)
To: Vivien Didelot, netdev; +Cc: andrew, john, davem
In-Reply-To: <8760g9kmy6.fsf@weeman.i-did-not-set--mail-host-address--so-tickle-me>
On 06/06/2017 11:09 AM, Vivien Didelot wrote:
> Florian Fainelli <f.fainelli@gmail.com> writes:
>
>> - phy_mode = of_get_phy_mode(ds->dst->cpu_dp->dn);
>> + phy_mode = of_get_phy_mode(ds->ports[QCA8K_CPU_PORT].dn);
>
> Is it necessary to use QCA8K_CPU_PORT?
>
>> +static inline struct dsa_port *dsa_ds_get_cpu_dp(struct dsa_switch *ds)
>> +{
>> + return &ds->ports[fls(ds->cpu_port_mask) - 1];
>> +}
>
> Wouldn't it be better to return the CPU port for a given port?
> Something like return ds->ports[port].cpu_dp, so that we ease the
> introduction of multiple CPU port a bit more?
ds->ports[port].cpu_dp only gets assigned at dsa_slave_create() time,
which is after ops->setup() has been called, hence this helper function
in case you need it earlier (e.g: like mv88e6060).
--
Florian
^ permalink raw reply
* Re: [PATCH net-next 0/5] net: dsa: Multi-CPU ground work
From: Vivien Didelot @ 2017-06-06 18:25 UTC (permalink / raw)
To: Florian Fainelli, netdev; +Cc: andrew, john, davem, Florian Fainelli
In-Reply-To: <20170606170202.21491-1-f.fainelli@gmail.com>
Hi Florian,
Florian Fainelli <f.fainelli@gmail.com> writes:
> This patch series prepares the ground for adding mutliple CPU port support to
multiple
> DSA, and starts by removing redundant pieces of information such as
> master_netdev which is cpu_dp->ethernet. Finally drivers are moved away from
cpu_dp->netdev
> directly accessing ds->dst->cpu_dp and use appropriate helper functions.
>
> Note that if you have Device Tree blobs/platform configurations that are
> currently listing multiple CPU ports, the proposed behavior in
> dsa_ds_get_cpu_dp() will be to return the last bit set in ds->cpu_port_mask.
>
> Future plans include:
> - making dst->cpu_dp a flexible data structure (array, list, you name it)
> - having the ability for drivers to return a default/preferred CPU port (if
> necessary)
The overall patchset looks good. I have questions for future work
though.
I am still not sure that we need CPU port references in
dsa_switch_tree. When device tree or pdata is parsed, we have allocated
dsa_switch and dsa_port structures. We should be able validate and
assign all ds->ports[x].cpu_dp, before setting up the switches and
creating the slave devices. What do you think?
Also I see dsa_ptr becoming a pointer to the assosicated dsa_port, and
dsa_port should contain the tagging ops for quick access. That is more
rigourous with the physical representation and much easier for
transparent multi-CPU port support.
Thanks,
Vivien
^ permalink raw reply
* Re: [PATCH v2 4/4] net: macb: Add hardware PTP support
From: Richard Cochran @ 2017-06-06 18:33 UTC (permalink / raw)
To: Rafal Ozieblo
Cc: David Miller, nicolas.ferre, netdev, linux-kernel, devicetree,
linux-arm-kernel, harini.katakam, andrei.pistirica
In-Reply-To: <1496413690-22826-1-git-send-email-rafalo@cadence.com>
On Fri, Jun 02, 2017 at 03:28:10PM +0100, Rafal Ozieblo wrote:
> +static s32 gem_get_ptp_max_adj(void)
> +{
> + return 64E6;
> +}
This is a floating point constant. Please use integer instead.
> +
> +static int gem_get_ts_info(struct net_device *dev,
> + struct ethtool_ts_info *info)
> +{
> + struct macb *bp = netdev_priv(dev);
> +
> + ethtool_op_get_ts_info(dev, info);
This default is misguided.
> + if ((bp->hw_dma_cap & HW_DMA_CAP_PTP) == 0)
> + return 0;
Try this:
if ((bp->hw_dma_cap & HW_DMA_CAP_PTP) == 0) {
ethtool_op_get_ts_info(dev, info);
return 0;
}
> + info->so_timestamping =
> + SOF_TIMESTAMPING_TX_SOFTWARE |
> + SOF_TIMESTAMPING_RX_SOFTWARE |
> + SOF_TIMESTAMPING_SOFTWARE |
> + SOF_TIMESTAMPING_TX_HARDWARE |
> + SOF_TIMESTAMPING_RX_HARDWARE |
> + SOF_TIMESTAMPING_RAW_HARDWARE;
> + info->tx_types =
> + (1 << HWTSTAMP_TX_ONESTEP_SYNC) |
> + (1 << HWTSTAMP_TX_OFF) |
> + (1 << HWTSTAMP_TX_ON);
> + info->rx_filters =
> + (1 << HWTSTAMP_FILTER_NONE) |
> + (1 << HWTSTAMP_FILTER_ALL);
> + info->phc_index = -1;
> +
> + if (bp->ptp_clock)
> + info->phc_index = ptp_clock_index(bp->ptp_clock);
Like this please:
info->phc_index = bp->ptp_clock ? ptp_clock_index(bp->ptp_clock) : -1;
> +
> + return 0;
> +}
> +
> +static struct macb_ptp_info gem_ptp_info = {
> + .ptp_init = gem_ptp_init,
> + .ptp_remove = gem_ptp_remove,
> + .get_ptp_max_adj = gem_get_ptp_max_adj,
> + .get_tsu_rate = gem_get_tsu_rate,
> + .get_ts_info = gem_get_ts_info,
> + .get_hwtst = gem_get_hwtst,
> + .set_hwtst = gem_set_hwtst,
> +};
> +#endif
> +
> static int macb_get_ts_info(struct net_device *netdev,
> struct ethtool_ts_info *info)
> {
> @@ -2636,12 +2707,16 @@ static void macb_configure_caps(struct macb *bp,
> dcfg = gem_readl(bp, DCFG2);
> if ((dcfg & (GEM_BIT(RX_PKT_BUFF) | GEM_BIT(TX_PKT_BUFF))) == 0)
> bp->caps |= MACB_CAPS_FIFO_MODE;
> - if (IS_ENABLED(CONFIG_MACB_USE_HWSTAMP) && gem_has_ptp(bp)) {
> +#ifdef CONFIG_MACB_USE_HWSTAMP
> + if (gem_has_ptp(bp)) {
> if (!GEM_BFEXT(TSU, gem_readl(bp, DCFG5)))
> pr_err("GEM doesn't support hardware ptp.\n");
> - else
> + else {
> bp->hw_dma_cap |= HW_DMA_CAP_PTP;
> + bp->ptp_info = &gem_ptp_info;
> + }
> }
> +#endif
> }
>
> dev_dbg(&bp->pdev->dev, "Cadence caps 0x%08x\n", bp->caps);
> @@ -3247,7 +3322,9 @@ static const struct macb_config np4_config = {
> };
>
> static const struct macb_config zynqmp_config = {
> - .caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE | MACB_CAPS_JUMBO,
> + .caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE |
> + MACB_CAPS_JUMBO |
> + MACB_CAPS_GEM_HAS_PTP,
> .dma_burst_length = 16,
> .clk_init = macb_clk_init,
> .init = macb_init,
> @@ -3281,7 +3358,9 @@ MODULE_DEVICE_TABLE(of, macb_dt_ids);
> #endif /* CONFIG_OF */
>
> static const struct macb_config default_gem_config = {
> - .caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE | MACB_CAPS_JUMBO,
> + .caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE |
> + MACB_CAPS_JUMBO |
> + MACB_CAPS_GEM_HAS_PTP,
> .dma_burst_length = 16,
> .clk_init = macb_clk_init,
> .init = macb_init,
> diff --git a/drivers/net/ethernet/cadence/macb_ptp.c b/drivers/net/ethernet/cadence/macb_ptp.c
> new file mode 100755
> index 0000000..d536970
> --- /dev/null
> +++ b/drivers/net/ethernet/cadence/macb_ptp.c
> @@ -0,0 +1,512 @@
> +/**
> + * 1588 PTP support for Cadence GEM device.
> + *
> + * Copyright (C) 2017 Cadence Design Systems - http://www.cadence.com
> + *
> + * Authors: Rafal Ozieblo <rafalo@cadence.com>
> + * Bartosz Folta <bfolta@cadence.com>
> + *
> + * This program is free software: you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 of
> + * the License as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
> + */
> +#include <linux/kernel.h>
> +#include <linux/types.h>
> +#include <linux/clk.h>
> +#include <linux/device.h>
> +#include <linux/etherdevice.h>
> +#include <linux/platform_device.h>
> +#include <linux/time64.h>
> +#include <linux/ptp_classify.h>
> +#include <linux/if_ether.h>
> +#include <linux/if_vlan.h>
> +#include <linux/net_tstamp.h>
> +#include <linux/circ_buf.h>
> +#include <linux/spinlock.h>
> +
> +#include "macb.h"
> +
> +#define GEM_PTP_TIMER_NAME "gem-ptp-timer"
> +
> +static struct macb_dma_desc_ptp *macb_ptp_desc(struct macb *bp,
> + struct macb_dma_desc *desc)
> +{
> + if (bp->hw_dma_cap == HW_DMA_CAP_PTP)
> + return (struct macb_dma_desc_ptp *)
> + ((u8 *)desc + sizeof(struct macb_dma_desc));
> + if (bp->hw_dma_cap == HW_DMA_CAP_64B_PTP)
> + return (struct macb_dma_desc_ptp *)
> + ((u8 *)desc + sizeof(struct macb_dma_desc)
> + + sizeof(struct macb_dma_desc_64));
> + return NULL;
> +}
> +
> +static int gem_tsu_get_time(struct ptp_clock_info *ptp, struct timespec64 *ts)
> +{
> + long first, second;
> + u32 secl, sech;
> + unsigned long flags;
> + struct macb *bp = container_of(ptp, struct macb, ptp_clock_info);
Please list locals in "upside down Christmas tree" style:
struct macb *bp = container_of(ptp, struct macb, ptp_clock_info);
unsigned long flags;
long first, second;
u32 secl, sech;
> + spin_lock_irqsave(&bp->tsu_clk_lock, flags);
> + first = gem_readl(bp, TN);
> + secl = gem_readl(bp, TSL);
> + sech = gem_readl(bp, TSH);
> + second = gem_readl(bp, TN);
> +
> + /* test for nsec rollover */
> + if (first > second) {
> + /* if so, use later read & re-read seconds
> + * (assume all done within 1s)
> + */
> + ts->tv_nsec = gem_readl(bp, TN);
> + secl = gem_readl(bp, TSL);
> + sech = gem_readl(bp, TSH);
> + } else {
> + ts->tv_nsec = first;
> + }
> +
> + spin_unlock_irqrestore(&bp->tsu_clk_lock, flags);
> + ts->tv_sec = (((u64)sech << GEM_TSL_SIZE) | secl)
> + & TSU_SEC_MAX_VAL;
> + return 0;
> +}
> +
> +static int gem_tsu_set_time(struct ptp_clock_info *ptp,
> + const struct timespec64 *ts)
> +{
> + u32 ns, sech, secl;
> + unsigned long flags;
> + struct macb *bp = container_of(ptp, struct macb, ptp_clock_info);
here too ...
> + secl = (u32)ts->tv_sec;
> + sech = (ts->tv_sec >> GEM_TSL_SIZE) & ((1 << GEM_TSH_SIZE) - 1);
> + ns = ts->tv_nsec;
> +
> + spin_lock_irqsave(&bp->tsu_clk_lock, flags);
> +
> + /* TSH doesn't latch the time and no atomicity! */
> + gem_writel(bp, TN, 0); /* clear to avoid overflow */
> + gem_writel(bp, TSH, sech);
> + /* write lower bits 2nd, for synchronized secs update */
> + gem_writel(bp, TSL, secl);
> + gem_writel(bp, TN, ns);
> +
> + spin_unlock_irqrestore(&bp->tsu_clk_lock, flags);
> +
> + return 0;
> +}
> +
> +static int gem_tsu_incr_set(struct macb *bp, struct tsu_incr *incr_spec)
> +{
> + unsigned long flags;
> +
> + /* tsu_timer_incr register must be written after
> + * the tsu_timer_incr_sub_ns register and the write operation
> + * will cause the value written to the tsu_timer_incr_sub_ns register
> + * to take effect.
> + */
> + spin_lock_irqsave(&bp->tsu_clk_lock, flags);
> + gem_writel(bp, TISUBN, GEM_BF(SUBNSINCR, incr_spec->sub_ns));
> + gem_writel(bp, TI, GEM_BF(NSINCR, incr_spec->ns));
> + spin_unlock_irqrestore(&bp->tsu_clk_lock, flags);
> +
> + return 0;
> +}
> +
> +static int gem_ptp_adjfine(struct ptp_clock_info *ptp, long scaled_ppm)
> +{
> + struct tsu_incr incr_spec;
> + struct macb *bp = container_of(ptp, struct macb, ptp_clock_info);
> + u64 adj;
> + u32 word;
> + bool neg_adj = false;
and here ...
> +
> + if (scaled_ppm < 0) {
> + neg_adj = true;
> + scaled_ppm = -scaled_ppm;
> + }
> +
> + /* Adjustment is relative to base frequency */
> + incr_spec.sub_ns = bp->tsu_incr.sub_ns;
> + incr_spec.ns = bp->tsu_incr.ns;
> +
> + /* scaling: unused(8bit) | ns(8bit) | fractions(16bit) */
> + word = ((u64)incr_spec.ns << GEM_SUBNSINCR_SIZE) + incr_spec.sub_ns;
> + adj = (u64)scaled_ppm * word;
> + /* Divide with rounding, equivalent to floating dividing:
> + * (temp / USEC_PER_SEC) + 0.5
> + */
> + adj += (USEC_PER_SEC >> 1);
> + adj >>= GEM_SUBNSINCR_SIZE; /* remove fractions */
> + adj = div_u64(adj, USEC_PER_SEC);
> + adj = neg_adj ? (word - adj) : (word + adj);
> +
> + incr_spec.ns = (adj >> GEM_SUBNSINCR_SIZE)
> + & ((1 << GEM_NSINCR_SIZE) - 1);
> + incr_spec.sub_ns = adj & ((1 << GEM_SUBNSINCR_SIZE) - 1);
> + gem_tsu_incr_set(bp, &incr_spec);
> + return 0;
> +}
> +
> +static int gem_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta)
> +{
> + struct timespec64 now, then = ns_to_timespec64(delta);
> + struct macb *bp = container_of(ptp, struct macb, ptp_clock_info);
> + u32 adj, sign = 0;
and here too. Please check the other functions...
> + if (delta < 0) {
> + sign = 1;
> + delta = -delta;
> + }
> +
> + if (delta > TSU_NSEC_MAX_VAL) {
> + gem_tsu_get_time(&bp->ptp_clock_info, &now);
> + if (sign)
> + now = timespec64_sub(now, then);
> + else
> + now = timespec64_add(now, then);
> +
> + gem_tsu_set_time(&bp->ptp_clock_info,
> + (const struct timespec64 *)&now);
> + } else {
> + adj = (sign << GEM_ADDSUB_OFFSET) | delta;
> +
> + gem_writel(bp, TA, adj);
> + }
> +
> + return 0;
> +}
> +
> +static int gem_ptp_enable(struct ptp_clock_info *ptp,
> + struct ptp_clock_request *rq, int on)
> +{
> + return -EOPNOTSUPP;
> +}
> +
> +static struct ptp_clock_info gem_ptp_caps_template = {
> + .owner = THIS_MODULE,
> + .name = GEM_PTP_TIMER_NAME,
> + .max_adj = 0,
> + .n_alarm = 0,
> + .n_ext_ts = 0,
> + .n_per_out = 0,
> + .n_pins = 0,
> + .pps = 1,
> + .adjfine = gem_ptp_adjfine,
> + .adjtime = gem_ptp_adjtime,
> + .gettime64 = gem_tsu_get_time,
> + .settime64 = gem_tsu_set_time,
> + .enable = gem_ptp_enable,
> +};
> +
> +static void gem_ptp_init_timer(struct macb *bp)
> +{
> + u32 rem = 0;
> +
> + bp->tsu_incr.ns = div_u64_rem(NSEC_PER_SEC, bp->tsu_rate, &rem);
> + if (rem) {
> + u64 adj = rem;
This variable belongs above, not in the if-block.
> + adj <<= GEM_SUBNSINCR_SIZE;
> + bp->tsu_incr.sub_ns = div_u64(adj, bp->tsu_rate);
> + } else {
> + bp->tsu_incr.sub_ns = 0;
> + }
> +}
> +
> +static void gem_ptp_init_tsu(struct macb *bp)
> +{
> + struct timespec64 ts;
> +
> + /* 1. get current system time */
> + ts = ns_to_timespec64(ktime_to_ns(ktime_get_real()));
> +
> + /* 2. set ptp timer */
> + gem_tsu_set_time(&bp->ptp_clock_info, &ts);
> +
> + /* 3. set PTP timer increment value to BASE_INCREMENT */
> + gem_tsu_incr_set(bp, &bp->tsu_incr);
> +
> + gem_writel(bp, TA, 0);
> +}
> +
> +static void gem_ptp_clear_timer(struct macb *bp)
> +{
> + bp->tsu_incr.ns = 0;
> + bp->tsu_incr.sub_ns = 0;
What is the point of this function?
> + gem_writel(bp, TISUBN, GEM_BF(SUBNSINCR, 0));
> + gem_writel(bp, TI, GEM_BF(NSINCR, 0));
> + gem_writel(bp, TA, 0);
> +}
> +
> +static int gem_hw_timestamp(struct macb *bp, u32 dma_desc_ts_1,
> + u32 dma_desc_ts_2, struct timespec64 *ts)
> +{
> + struct timespec64 tsu;
> +
> + ts->tv_sec = (GEM_BFEXT(DMA_SECH, dma_desc_ts_2) << GEM_DMA_SECL_SIZE) |
> + GEM_BFEXT(DMA_SECL, dma_desc_ts_1);
> + ts->tv_nsec = GEM_BFEXT(DMA_NSEC, dma_desc_ts_1);
> +
> + /* TSU overlapping workaround
> + * The timestamp only contains lower few bits of seconds,
> + * so add value from 1588 timer
> + */
> + gem_tsu_get_time(&bp->ptp_clock_info, &tsu);
> +
> + /* If the top bit is set in the timestamp,
> + * but not in 1588 timer, it has rolled over,
> + * so subtract max size
> + */
> + if ((ts->tv_sec & (GEM_DMA_SEC_TOP >> 1)) &&
> + !(tsu.tv_sec & (GEM_DMA_SEC_TOP >> 1)))
> + ts->tv_sec -= GEM_DMA_SEC_TOP;
> +
> + ts->tv_sec += ((~GEM_DMA_SEC_MASK) & tsu.tv_sec);
> +
> + return 0;
> +}
> +
> +void gem_ptp_rxstamp(struct macb *bp, struct sk_buff *skb,
> + struct macb_dma_desc *desc)
> +{
> + struct timespec64 ts;
> + struct skb_shared_hwtstamps *shhwtstamps = skb_hwtstamps(skb);
> + struct macb_dma_desc_ptp *desc_ptp;
> +
> + if (GEM_BFEXT(DMA_RXVALID, desc->addr)) {
> + desc_ptp = macb_ptp_desc(bp, desc);
> + gem_hw_timestamp(bp, desc_ptp->ts_1, desc_ptp->ts_2, &ts);
> + memset(shhwtstamps, 0, sizeof(struct skb_shared_hwtstamps));
> + shhwtstamps->hwtstamp = ktime_set(ts.tv_sec, ts.tv_nsec);
> + }
> +}
> +
> +static void gem_tstamp_tx(struct macb *bp, struct sk_buff *skb,
> + struct macb_dma_desc_ptp *desc_ptp)
> +{
> + struct skb_shared_hwtstamps shhwtstamps;
> + struct timespec64 ts;
> +
> + gem_hw_timestamp(bp, desc_ptp->ts_1, desc_ptp->ts_2, &ts);
> + memset(&shhwtstamps, 0, sizeof(shhwtstamps));
> + shhwtstamps.hwtstamp = ktime_set(ts.tv_sec, ts.tv_nsec);
> + skb_tstamp_tx(skb, &shhwtstamps);
> +}
> +
> +int gem_ptp_txstamp(struct macb_queue *queue, struct sk_buff *skb,
> + struct macb_dma_desc *desc)
> +{
> + struct gem_tx_ts *tx_timestamp;
> + struct macb_dma_desc_ptp *desc_ptp;
> + unsigned long head = queue->tx_ts_head;
> + unsigned long tail = READ_ONCE(queue->tx_ts_tail);
> +
> + if (!GEM_BFEXT(DMA_TXVALID, desc->ctrl))
> + return -EINVAL;
> +
> + if (CIRC_SPACE(head, tail, PTP_TS_BUFFER_SIZE) == 0)
> + return -ENOMEM;
> +
> + skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
> + desc_ptp = macb_ptp_desc(queue->bp, desc);
> + tx_timestamp = &queue->tx_timestamps[head];
> + tx_timestamp->skb = skb;
> + tx_timestamp->desc_ptp.ts_1 = desc_ptp->ts_1;
> + tx_timestamp->desc_ptp.ts_2 = desc_ptp->ts_2;
> + /* move head */
> + smp_store_release(&queue->tx_ts_head,
> + (head + 1) & (PTP_TS_BUFFER_SIZE - 1));
> +
> + schedule_work(&queue->tx_ts_task);
Since the time stamp is in the buffer descriptor, why delay the
delivery via the work item?
> + return 0;
> +}
> +
> +static void gem_tx_timestamp_flush(struct work_struct *work)
> +{
> + struct macb_queue *queue =
> + container_of(work, struct macb_queue, tx_ts_task);
> + struct gem_tx_ts *tx_ts;
> + unsigned long head, tail;
> +
> + /* take current head */
> + head = smp_load_acquire(&queue->tx_ts_head);
> + tail = queue->tx_ts_tail;
> +
> + while (CIRC_CNT(head, tail, PTP_TS_BUFFER_SIZE)) {
> + tx_ts = &queue->tx_timestamps[tail];
> + gem_tstamp_tx(queue->bp, tx_ts->skb, &tx_ts->desc_ptp);
> + /* cleanup */
> + dev_kfree_skb_any(tx_ts->skb);
> + /* remove old tail */
> + smp_store_release(&queue->tx_ts_tail,
> + (tail + 1) & (PTP_TS_BUFFER_SIZE - 1));
> + tail = queue->tx_ts_tail;
> + }
> +}
> +
> +void gem_ptp_init(struct net_device *dev)
> +{
> + struct macb *bp = netdev_priv(dev);
> + unsigned int q;
> + struct macb_queue *queue;
> +
> + bp->ptp_clock_info = gem_ptp_caps_template;
> +
> + /* nominal frequency and maximum adjustment in ppb */
> + bp->tsu_rate = bp->ptp_info->get_tsu_rate(bp);
> + bp->ptp_clock_info.max_adj = bp->ptp_info->get_ptp_max_adj();
> + gem_ptp_init_timer(bp);
> + bp->ptp_clock = ptp_clock_register(&bp->ptp_clock_info, &dev->dev);
> + if (IS_ERR(&bp->ptp_clock)) {
> + bp->ptp_clock = NULL;
> + pr_err("ptp clock register failed\n");
> + return;
> + }
ptp_clock_register() can also return NULL, and so you can avoid the
following code in that case, right?
> +
> + spin_lock_init(&bp->tsu_clk_lock);
> + for (q = 0, queue = bp->queues; q < bp->num_queues; ++q, ++queue) {
> + queue->tx_ts_head = 0;
> + queue->tx_ts_tail = 0;
> + INIT_WORK(&queue->tx_ts_task, gem_tx_timestamp_flush);
> + }
> +
> + gem_ptp_init_tsu(bp);
> +
> + dev_info(&bp->pdev->dev, "%s ptp clock registered.\n",
> + GEM_PTP_TIMER_NAME);
> +}
> +
> +void gem_ptp_remove(struct net_device *ndev)
> +{
> + struct macb *bp = netdev_priv(ndev);
> +
> + if (bp->ptp_clock)
> + ptp_clock_unregister(bp->ptp_clock);
> +
> + gem_ptp_clear_timer(bp);
Why is this 'clear' needed?
> + dev_info(&bp->pdev->dev, "%s ptp clock unregistered.\n",
> + GEM_PTP_TIMER_NAME);
> +}
Thanks,
Richard
^ permalink raw reply
* Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
From: Alexei Starovoitov @ 2017-06-06 18:34 UTC (permalink / raw)
To: David Miller
Cc: jgunthorpe, ilant, saeedm, dledford, netdev, linux-rdma,
jsorensen, andy.shevchenko, linux-fpga, atull, yi1.li, borisp
In-Reply-To: <20170606.134726.416045938000214127.davem@davemloft.net>
On Tue, Jun 06, 2017 at 01:47:26PM -0400, David Miller wrote:
> From: Alexei Starovoitov <alexei.starovoitov@gmail.com>
> Date: Tue, 6 Jun 2017 10:42:35 -0700
>
> > so it's like rdma, but without using kernel rdma stack?
>
> No sockets here, just transformation rules. It's like offloading
> a complex TC rule to hardware version of that transformation.
>
> Yes, there is state, but I argue that it is no different than TC
> offloading rules. What if TC had "hash" and "crypt" operations
> and we attached them to appropriate u32 matches? You wouldn't
> be able to tell the difference.
there is huge difference in underlying hw.
fpga is a separate device with its own phy and mac layers, its
own queues, packet parsing and rdma logic.
Where as tc offload is happening within the same hw queues/memory/stats
management logic. My understanding that when I do 'ethtool -L' to
change number of queues or 'ethtool -G' it changes the memory layout
that tc offload is operating on as well.
When I do 'ethtool -S' it shows me the stats for the device
that tc offload rules are integral part of.
Whereas fpga is a different physical device with its own
buffers and such. We can add 'ethtool -G_fpga, -L_fpga', etc
but this type of discussion needs to happen _before_ the whole
thing is merged. It will never happen after the fact.
Just look at mlx responses, they still don't acknowledge the issue
and instead pushing for ipsec, tls (in other words: new features)
instead of addressing production issues that are obviously
not glamorous to work on and fix.
> I think you are way over-obsessed with this FPGA offload thing,
> quite frankly.
if we didn't have issues with eswitch that drops packets and
we don't even know how many, I wouldn't be complaining.
There is a discussion going on to add few counters for
eswitch visibility, but it's taking forever and it's not at
the point of exposing eswitch as a kernel object.
Why? because it's hard to refactor it now into something like devlink
or whatever new abstraction that would be needed.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox