* Re: [PATCH v2 1/2] mac80211: Add rcu read side critical sections
From: Ville Syrjälä @ 2017-09-20 12:11 UTC (permalink / raw)
To: Johannes Berg
Cc: linux-wireless-u79uwXL29TY76Z2rM5mHXA, David S. Miller,
netdev-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1505903964.3026.14.camel-cdvu00un1VgdHxzADdlk8Q@public.gmane.org>
On Wed, Sep 20, 2017 at 12:39:24PM +0200, Johannes Berg wrote:
> On Wed, 2017-09-20 at 13:11 +0300, Ville Syrjala wrote:
>
> > --- a/net/mac80211/tx.c
> > +++ b/net/mac80211/tx.c
> > @@ -1770,15 +1770,21 @@ bool ieee80211_tx_prepare_skb(struct ieee80211_hw *hw,
> > struct ieee80211_tx_data tx;
> > struct sk_buff *skb2;
> >
> > - if (ieee80211_tx_prepare(sdata, &tx, NULL, skb) == TX_DROP)
> > + rcu_read_lock();
>
> The documentation says:
>
> /**
> * ieee80211_tx_prepare_skb - prepare an 802.11 skb for transmission
> * @hw: pointer as obtained from ieee80211_alloc_hw()
> * @vif: virtual interface
> * @skb: frame to be sent from within the driver
> * @band: the band to transmit on
> * @sta: optional pointer to get the station to send the frame to
> *
> * Note: must be called under RCU lock
> */
>
> You can't even argue that it should be the function itself doing it,
> because the (admittedly optional) sta pointer would otherwise not have
> proper protection after you leave the function ... You can't pass out a
> sta pointer that's RCU protected.
Yeah, I suppose that would need rcu_handoff+some other mechanism to
make sure it stays around after that.
>
> Side note: Perhaps some annotation should be there? not sure it's
> possible - would have to be something like
> struct ieee80211_sta * __rcu *sta;
>
> I guess since the outer pointer isn't protected, only the inner ...
I think just the fact that even the pointers in ieee80211_tx_data don't
have the __rcu annotation makes it rather hard to see what is really rcu
protected and what isn't. If every user of those pointers would have to
do the rcu_dereference() things would be rather more explicit.
> Therefore, this patch is wrong.
OK, so the problem is in ath9k then.
> I actually think the same is true for ieee80211_tx_dequeue(), but I'm
> less sure about it - the sta pointer there clearly is somehow safely
> passed in (even if it's w/o RCU, the driver can potentially make that
> safe), but the key pointer seems unsafe in this case (as well) if
> there's no outer RCU protection.
Well, I think this is as far as I want to dig into the matter. I can
respin the patch once more with just tx_dequeue() fix in there, if you
want (not sure you do if you think it's wrong as well). After that I'll
leave it to someone who actually knows what they're doing with mac80211 ;)
--
Ville Syrjälä
Intel OTC
^ permalink raw reply
* [PATCH net-next 10/10] net/smc: no close wait in case of process shut down
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
Usually socket closing is delayed if there is still data available in
the send buffer to be transmitted. If a process is killed, the delay
should be avoided.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_close.c | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c
index 5201bc103bd8..f0d16fb825f7 100644
--- a/net/smc/smc_close.c
+++ b/net/smc/smc_close.c
@@ -174,15 +174,15 @@ int smc_close_active(struct smc_sock *smc)
{
struct smc_cdc_conn_state_flags *txflags =
&smc->conn.local_tx_ctrl.conn_state_flags;
- long timeout = SMC_MAX_STREAM_WAIT_TIMEOUT;
struct smc_connection *conn = &smc->conn;
struct sock *sk = &smc->sk;
int old_state;
+ long timeout;
int rc = 0;
- if (sock_flag(sk, SOCK_LINGER) &&
- !(current->flags & PF_EXITING))
- timeout = sk->sk_lingertime;
+ timeout = current->flags & PF_EXITING ?
+ 0 : sock_flag(sk, SOCK_LINGER) ?
+ sk->sk_lingertime : SMC_MAX_STREAM_WAIT_TIMEOUT;
again:
old_state = sk->sk_state;
@@ -413,13 +413,14 @@ void smc_close_sock_put_work(struct work_struct *work)
int smc_close_shutdown_write(struct smc_sock *smc)
{
struct smc_connection *conn = &smc->conn;
- long timeout = SMC_MAX_STREAM_WAIT_TIMEOUT;
struct sock *sk = &smc->sk;
int old_state;
+ long timeout;
int rc = 0;
- if (sock_flag(sk, SOCK_LINGER))
- timeout = sk->sk_lingertime;
+ timeout = current->flags & PF_EXITING ?
+ 0 : sock_flag(sk, SOCK_LINGER) ?
+ sk->sk_lingertime : SMC_MAX_STREAM_WAIT_TIMEOUT;
again:
old_state = sk->sk_state;
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 09/10] net/smc: parameter cleanup in smc_cdc_get_free_slot()
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
Use the smc_connection as first parameter with smc_cdc_get_free_slot().
This is just a small code cleanup, no functional change.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_cdc.c | 7 ++++---
net/smc/smc_cdc.h | 3 ++-
net/smc/smc_tx.c | 6 ++----
3 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/net/smc/smc_cdc.c b/net/smc/smc_cdc.c
index a7294edbc221..5ef97e5a5f78 100644
--- a/net/smc/smc_cdc.c
+++ b/net/smc/smc_cdc.c
@@ -62,10 +62,12 @@ static void smc_cdc_tx_handler(struct smc_wr_tx_pend_priv *pnd_snd,
bh_unlock_sock(&smc->sk);
}
-int smc_cdc_get_free_slot(struct smc_link *link,
+int smc_cdc_get_free_slot(struct smc_connection *conn,
struct smc_wr_buf **wr_buf,
struct smc_cdc_tx_pend **pend)
{
+ struct smc_link *link = &conn->lgr->lnk[SMC_SINGLE_LINK];
+
return smc_wr_tx_get_free_slot(link, smc_cdc_tx_handler, wr_buf,
(struct smc_wr_tx_pend_priv **)pend);
}
@@ -118,8 +120,7 @@ int smc_cdc_get_slot_and_msg_send(struct smc_connection *conn)
struct smc_wr_buf *wr_buf;
int rc;
- rc = smc_cdc_get_free_slot(&conn->lgr->lnk[SMC_SINGLE_LINK], &wr_buf,
- &pend);
+ rc = smc_cdc_get_free_slot(conn, &wr_buf, &pend);
if (rc)
return rc;
diff --git a/net/smc/smc_cdc.h b/net/smc/smc_cdc.h
index 8e1d76f26007..56f883d1159c 100644
--- a/net/smc/smc_cdc.h
+++ b/net/smc/smc_cdc.h
@@ -206,7 +206,8 @@ static inline void smc_cdc_msg_to_host(struct smc_host_cdc_msg *local,
struct smc_cdc_tx_pend;
-int smc_cdc_get_free_slot(struct smc_link *link, struct smc_wr_buf **wr_buf,
+int smc_cdc_get_free_slot(struct smc_connection *conn,
+ struct smc_wr_buf **wr_buf,
struct smc_cdc_tx_pend **pend);
void smc_cdc_tx_dismiss_slots(struct smc_connection *conn);
int smc_cdc_msg_send(struct smc_connection *conn, struct smc_wr_buf *wr_buf,
diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c
index 3866573288dd..ec49bc3c3283 100644
--- a/net/smc/smc_tx.c
+++ b/net/smc/smc_tx.c
@@ -396,8 +396,7 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn)
int rc;
spin_lock_bh(&conn->send_lock);
- rc = smc_cdc_get_free_slot(&conn->lgr->lnk[SMC_SINGLE_LINK], &wr_buf,
- &pend);
+ rc = smc_cdc_get_free_slot(conn, &wr_buf, &pend);
if (rc < 0) {
if (rc == -EBUSY) {
struct smc_sock *smc =
@@ -466,8 +465,7 @@ void smc_tx_consumer_update(struct smc_connection *conn)
((to_confirm > conn->rmbe_update_limit) &&
((to_confirm > (conn->rmbe_size / 2)) ||
conn->local_rx_ctrl.prod_flags.write_blocked))) {
- rc = smc_cdc_get_free_slot(&conn->lgr->lnk[SMC_SINGLE_LINK],
- &wr_buf, &pend);
+ rc = smc_cdc_get_free_slot(conn, &wr_buf, &pend);
if (!rc)
rc = smc_cdc_msg_send(conn, wr_buf, pend);
if (rc < 0) {
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 08/10] net/smc: introduce a delay
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
The number of outstanding work requests is limited. If all work
requests are in use, tx processing is postponed to another scheduling
of the tx worker. Switch to a delayed worker to have a gap for tx
completion queue events before the next retry.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc.h | 2 +-
net/smc/smc_close.c | 12 +++++++-----
net/smc/smc_tx.c | 12 ++++++++----
3 files changed, 16 insertions(+), 10 deletions(-)
diff --git a/net/smc/smc.h b/net/smc/smc.h
index 6e44313e4467..0ccd6fa387ad 100644
--- a/net/smc/smc.h
+++ b/net/smc/smc.h
@@ -149,7 +149,7 @@ struct smc_connection {
atomic_t sndbuf_space; /* remaining space in sndbuf */
u16 tx_cdc_seq; /* sequence # for CDC send */
spinlock_t send_lock; /* protect wr_sends */
- struct work_struct tx_work; /* retry of smc_cdc_msg_send */
+ struct delayed_work tx_work; /* retry of smc_cdc_msg_send */
struct smc_host_cdc_msg local_rx_ctrl; /* filled during event_handl.
* .prod cf. TCP rcv_nxt
diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c
index 3c2e166b5d22..5201bc103bd8 100644
--- a/net/smc/smc_close.c
+++ b/net/smc/smc_close.c
@@ -208,7 +208,7 @@ int smc_close_active(struct smc_sock *smc)
case SMC_ACTIVE:
smc_close_stream_wait(smc, timeout);
release_sock(sk);
- cancel_work_sync(&conn->tx_work);
+ cancel_delayed_work_sync(&conn->tx_work);
lock_sock(sk);
if (sk->sk_state == SMC_ACTIVE) {
/* send close request */
@@ -234,7 +234,7 @@ int smc_close_active(struct smc_sock *smc)
if (!smc_cdc_rxed_any_close(conn))
smc_close_stream_wait(smc, timeout);
release_sock(sk);
- cancel_work_sync(&conn->tx_work);
+ cancel_delayed_work_sync(&conn->tx_work);
lock_sock(sk);
if (sk->sk_err != ECONNABORTED) {
/* confirm close from peer */
@@ -263,7 +263,9 @@ int smc_close_active(struct smc_sock *smc)
/* peer sending PeerConnectionClosed will cause transition */
break;
case SMC_PROCESSABORT:
- cancel_work_sync(&conn->tx_work);
+ release_sock(sk);
+ cancel_delayed_work_sync(&conn->tx_work);
+ lock_sock(sk);
smc_close_abort(conn);
sk->sk_state = SMC_CLOSED;
smc_close_wait_tx_pends(smc);
@@ -425,7 +427,7 @@ int smc_close_shutdown_write(struct smc_sock *smc)
case SMC_ACTIVE:
smc_close_stream_wait(smc, timeout);
release_sock(sk);
- cancel_work_sync(&conn->tx_work);
+ cancel_delayed_work_sync(&conn->tx_work);
lock_sock(sk);
/* send close wr request */
rc = smc_close_wr(conn);
@@ -439,7 +441,7 @@ int smc_close_shutdown_write(struct smc_sock *smc)
if (!smc_cdc_rxed_any_close(conn))
smc_close_stream_wait(smc, timeout);
release_sock(sk);
- cancel_work_sync(&conn->tx_work);
+ cancel_delayed_work_sync(&conn->tx_work);
lock_sock(sk);
/* confirm close from peer */
rc = smc_close_wr(conn);
diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c
index 3c656beb8820..3866573288dd 100644
--- a/net/smc/smc_tx.c
+++ b/net/smc/smc_tx.c
@@ -24,6 +24,8 @@
#include "smc_cdc.h"
#include "smc_tx.h"
+#define SMC_TX_WORK_DELAY HZ
+
/***************************** sndbuf producer *******************************/
/* callback implementation for sk.sk_write_space()
@@ -406,7 +408,8 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn)
goto out_unlock;
}
rc = 0;
- schedule_work(&conn->tx_work);
+ schedule_delayed_work(&conn->tx_work,
+ SMC_TX_WORK_DELAY);
}
goto out_unlock;
}
@@ -430,7 +433,7 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn)
*/
static void smc_tx_work(struct work_struct *work)
{
- struct smc_connection *conn = container_of(work,
+ struct smc_connection *conn = container_of(to_delayed_work(work),
struct smc_connection,
tx_work);
struct smc_sock *smc = container_of(conn, struct smc_sock, conn);
@@ -468,7 +471,8 @@ void smc_tx_consumer_update(struct smc_connection *conn)
if (!rc)
rc = smc_cdc_msg_send(conn, wr_buf, pend);
if (rc < 0) {
- schedule_work(&conn->tx_work);
+ schedule_delayed_work(&conn->tx_work,
+ SMC_TX_WORK_DELAY);
return;
}
smc_curs_write(&conn->rx_curs_confirmed,
@@ -487,6 +491,6 @@ void smc_tx_consumer_update(struct smc_connection *conn)
void smc_tx_init(struct smc_sock *smc)
{
smc->sk.sk_write_space = smc_tx_write_space;
- INIT_WORK(&smc->conn.tx_work, smc_tx_work);
+ INIT_DELAYED_WORK(&smc->conn.tx_work, smc_tx_work);
spin_lock_init(&smc->conn.send_lock);
}
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 05/10] net/smc: adapt send request completion notification
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
The solicited flag is meaningful for the receive completion queue.
Ask for next work completion of any type on the send queue.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_wr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c
index ab56bda66783..525d91e0d57e 100644
--- a/net/smc/smc_wr.c
+++ b/net/smc/smc_wr.c
@@ -244,7 +244,7 @@ int smc_wr_tx_send(struct smc_link *link, struct smc_wr_tx_pend_priv *priv)
int rc;
ib_req_notify_cq(link->smcibdev->roce_cq_send,
- IB_CQ_SOLICITED_MASK | IB_CQ_REPORT_MISSED_EVENTS);
+ IB_CQ_NEXT_COMP | IB_CQ_REPORT_MISSED_EVENTS);
pend = container_of(priv, struct smc_wr_tx_pend, priv);
rc = ib_post_send(link->roce_qp, &link->wr_tx_ibs[pend->idx],
&failed_wr);
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 07/10] net/smc: terminate link group if out-of-sync is received
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
An out-of-sync condition can just be detected by the client.
If the server receives a CLC DECLINE message indicating an out-of-sync
condition for the link groups, the server must clean up the out-of-sync
link group.
There is no need for an extra third parameter in smc_clc_send_decline().
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/af_smc.c | 6 ++----
net/smc/smc_clc.c | 10 +++++-----
net/smc/smc_clc.h | 3 +--
3 files changed, 8 insertions(+), 11 deletions(-)
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 2e8d2dabac0c..745f145d4c4d 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -513,7 +513,7 @@ static int smc_connect_rdma(struct smc_sock *smc)
/* RDMA setup failed, switch back to TCP */
smc->use_fallback = true;
if (reason_code && (reason_code != SMC_CLC_DECL_REPLY)) {
- rc = smc_clc_send_decline(smc, reason_code, 0);
+ rc = smc_clc_send_decline(smc, reason_code);
if (rc < sizeof(struct smc_clc_msg_decline))
goto out_err;
}
@@ -808,8 +808,6 @@ static void smc_listen_work(struct work_struct *work)
rc = local_contact;
if (rc == -ENOMEM)
reason_code = SMC_CLC_DECL_MEM;/* insufficient memory*/
- else if (rc == -ENOLINK)
- reason_code = SMC_CLC_DECL_SYNCERR; /* synchr. error */
goto decline_rdma;
}
link = &new_smc->conn.lgr->lnk[SMC_SINGLE_LINK];
@@ -903,7 +901,7 @@ static void smc_listen_work(struct work_struct *work)
smc_conn_free(&new_smc->conn);
new_smc->use_fallback = true;
if (reason_code && (reason_code != SMC_CLC_DECL_REPLY)) {
- rc = smc_clc_send_decline(new_smc, reason_code, 0);
+ rc = smc_clc_send_decline(new_smc, reason_code);
if (rc < sizeof(struct smc_clc_msg_decline))
goto out_err;
}
diff --git a/net/smc/smc_clc.c b/net/smc/smc_clc.c
index 3934913ab835..b7dd2743fb5c 100644
--- a/net/smc/smc_clc.c
+++ b/net/smc/smc_clc.c
@@ -95,9 +95,10 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
}
if (clcm->type == SMC_CLC_DECLINE) {
reason_code = SMC_CLC_DECL_REPLY;
- if (ntohl(((struct smc_clc_msg_decline *)buf)->peer_diagnosis)
- == SMC_CLC_DECL_SYNCERR)
+ if (((struct smc_clc_msg_decline *)buf)->hdr.flag) {
smc->conn.lgr->sync_err = true;
+ smc_lgr_terminate(smc->conn.lgr);
+ }
}
out:
@@ -105,8 +106,7 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
}
/* send CLC DECLINE message across internal TCP socket */
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
- u8 out_of_sync)
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
{
struct smc_clc_msg_decline dclc;
struct msghdr msg;
@@ -118,7 +118,7 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
dclc.hdr.type = SMC_CLC_DECLINE;
dclc.hdr.length = htons(sizeof(struct smc_clc_msg_decline));
dclc.hdr.version = SMC_CLC_V1;
- dclc.hdr.flag = out_of_sync ? 1 : 0;
+ dclc.hdr.flag = (peer_diag_info == SMC_CLC_DECL_SYNCERR) ? 1 : 0;
memcpy(dclc.id_for_peer, local_systemid, sizeof(local_systemid));
dclc.peer_diagnosis = htonl(peer_diag_info);
memcpy(dclc.trl.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER));
diff --git a/net/smc/smc_clc.h b/net/smc/smc_clc.h
index 13db8ce177c9..1c55414041d4 100644
--- a/net/smc/smc_clc.h
+++ b/net/smc/smc_clc.h
@@ -106,8 +106,7 @@ struct smc_ib_device;
int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
u8 expected_type);
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
- u8 out_of_sync);
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info);
int smc_clc_send_proposal(struct smc_sock *smc, struct smc_ib_device *smcibdev,
u8 ibport);
int smc_clc_send_confirm(struct smc_sock *smc);
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 06/10] net/smc: longer delay for client link group removal
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
Client link group creation always follows the server linkgroup creation.
If peer creates a new server link group, client has to create a new
client link group. If peer reuses a server link group for a new
connection, client has to reuse its client link group as well. This
patch introduces a longer delay for client link group removal to make
sure this link group still exists, once the peer decides to reuse a
server link group. This avoids out-of-sync conditions for link groups.
If already scheduled, modify the delay.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_core.c | 16 ++++++++++++----
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c
index 1a16d51e2330..20b66e79c5d6 100644
--- a/net/smc/smc_core.c
+++ b/net/smc/smc_core.c
@@ -25,8 +25,9 @@
#include "smc_cdc.h"
#include "smc_close.h"
-#define SMC_LGR_NUM_INCR 256
-#define SMC_LGR_FREE_DELAY (600 * HZ)
+#define SMC_LGR_NUM_INCR 256
+#define SMC_LGR_FREE_DELAY_SERV (600 * HZ)
+#define SMC_LGR_FREE_DELAY_CLNT (SMC_LGR_FREE_DELAY_SERV + 10)
static u32 smc_lgr_num; /* unique link group number */
@@ -107,8 +108,15 @@ static void smc_lgr_unregister_conn(struct smc_connection *conn)
__smc_lgr_unregister_conn(conn);
}
write_unlock_bh(&lgr->conns_lock);
- if (reduced && !lgr->conns_num)
- schedule_delayed_work(&lgr->free_work, SMC_LGR_FREE_DELAY);
+ if (!reduced || lgr->conns_num)
+ return;
+ /* client link group creation always follows the server link group
+ * creation. For client use a somewhat higher removal delay time,
+ * otherwise there is a risk of out-of-sync link groups.
+ */
+ mod_delayed_work(system_wq, &lgr->free_work,
+ lgr->role == SMC_CLNT ? SMC_LGR_FREE_DELAY_CLNT :
+ SMC_LGR_FREE_DELAY_SERV);
}
static void smc_lgr_free_work(struct work_struct *work)
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 04/10] net/smc: adjust net_device refcount
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
smc_pnet_fill_entry() uses dev_get_by_name() adding a refcount to ndev.
The following smc_pnet_enter() has to reduce the refcount if the entry
to be added exists already in the pnet table.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_pnet.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/net/smc/smc_pnet.c b/net/smc/smc_pnet.c
index 78f7af28ae4f..31f8453c25c5 100644
--- a/net/smc/smc_pnet.c
+++ b/net/smc/smc_pnet.c
@@ -181,8 +181,10 @@ static int smc_pnet_enter(struct smc_pnetentry *new_pnetelem)
sizeof(new_pnetelem->ndev->name)) ||
smc_pnet_same_ibname(pnetelem,
new_pnetelem->smcibdev->ibdev->name,
- new_pnetelem->ib_port))
+ new_pnetelem->ib_port)) {
+ dev_put(pnetelem->ndev);
goto found;
+ }
}
list_add_tail(&new_pnetelem->list, &smc_pnettable.pnetlist);
rc = 0;
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 02/10] net/smc: add receive timeout check
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
From: Hans Wippel <hwippel@linux.vnet.ibm.com>
The SMC receive function currently lacks a timeout check under the
condition that no data were received and no data are available. This
patch adds such a check.
Signed-off-by: Hans Wippel <hwippel@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_rx.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/net/smc/smc_rx.c b/net/smc/smc_rx.c
index b17a333e9bb0..3e631ae4b6b6 100644
--- a/net/smc/smc_rx.c
+++ b/net/smc/smc_rx.c
@@ -148,6 +148,8 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, size_t len,
read_done = sock_intr_errno(timeo);
break;
}
+ if (!timeo)
+ return -EAGAIN;
}
if (!atomic_read(&conn->bytes_to_rcv)) {
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 01/10] net/smc: add missing dev_put
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem
Cc: netdev, linux-rdma, linux-s390, jwi, schwidefsky, heiko.carstens,
raspl, ubraun
In-Reply-To: <20170920115813.63745-1-ubraun@linux.vnet.ibm.com>
From: Hans Wippel <hwippel@linux.vnet.ibm.com>
In the infiniband part, SMC currently uses get_netdev which calls
dev_hold on the returned net device. However, the SMC code never calls
dev_put on that net device resulting in a wrong reference count.
This patch adds a dev_put after the usage of the net device to fix the
issue.
Signed-off-by: Hans Wippel <hwippel@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
net/smc/smc_ib.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c
index 547e0e113b17..0b5852299158 100644
--- a/net/smc/smc_ib.c
+++ b/net/smc/smc_ib.c
@@ -380,6 +380,7 @@ static int smc_ib_fill_gid_and_mac(struct smc_ib_device *smcibdev, u8 ibport)
ndev = smcibdev->ibdev->get_netdev(smcibdev->ibdev, ibport);
if (ndev) {
memcpy(&smcibdev->mac, ndev->dev_addr, ETH_ALEN);
+ dev_put(ndev);
} else if (!rc) {
memcpy(&smcibdev->mac[ibport - 1][0],
&smcibdev->gid[ibport - 1].raw[8], 3);
--
2.13.5
^ permalink raw reply related
* [PATCH net-next 03/10] net/smc: take RCU read lock for routing cache lookup
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170920115813.63745-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
smc_netinfo_by_tcpsk() looks up the routing cache. Such a lookup requires
protection by an RCU read lock.
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/af_smc.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 8c6d24b2995d..2e8d2dabac0c 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -282,6 +282,7 @@ int smc_netinfo_by_tcpsk(struct socket *clcsock,
__be32 *subnet, u8 *prefix_len)
{
struct dst_entry *dst = sk_dst_get(clcsock->sk);
+ struct in_device *in_dev;
struct sockaddr_in addr;
int rc = -ENOENT;
int len;
@@ -298,14 +299,17 @@ int smc_netinfo_by_tcpsk(struct socket *clcsock,
/* get address to which the internal TCP socket is bound */
kernel_getsockname(clcsock, (struct sockaddr *)&addr, &len);
/* analyze IPv4 specific data of net_device belonging to TCP socket */
- for_ifa(dst->dev->ip_ptr) {
- if (ifa->ifa_address != addr.sin_addr.s_addr)
+ rcu_read_lock();
+ in_dev = __in_dev_get_rcu(dst->dev);
+ for_ifa(in_dev) {
+ if (!inet_ifa_match(addr.sin_addr.s_addr, ifa))
continue;
*prefix_len = inet_mask_len(ifa->ifa_mask);
*subnet = ifa->ifa_address & ifa->ifa_mask;
rc = 0;
break;
- } endfor_ifa(dst->dev->ip_ptr);
+ } endfor_ifa(in_dev);
+ rcu_read_unlock();
out_rel:
dst_release(dst);
--
2.13.5
--
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 related
* [PATCH net-next 00/10] net/smc: updates 2017-09-20
From: Ursula Braun @ 2017-09-20 11:58 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
Hi Dave,
here is a collection of small smc-patches built for net-next improving
the smc code in different areas.
Thanks,
Ursula
Hans Wippel (2):
net/smc: add missing dev_put
net/smc: add receive timeout check
Ursula Braun (8):
net/smc: take RCU read lock for routing cache lookup
net/smc: adjust net_device refcount
net/smc: adapt send request completion notification
net/smc: longer delay for client link group removal
net/smc: terminate link group if out-of-sync is received
net/smc: introduce a delay
net/smc: parameter cleanup in smc_cdc_get_free_slot()
net/smc: no close wait in case of process shut down
net/smc/af_smc.c | 16 +++++++++-------
net/smc/smc.h | 2 +-
net/smc/smc_cdc.c | 7 ++++---
net/smc/smc_cdc.h | 3 ++-
net/smc/smc_clc.c | 10 +++++-----
net/smc/smc_clc.h | 3 +--
net/smc/smc_close.c | 27 +++++++++++++++------------
net/smc/smc_core.c | 16 ++++++++++++----
net/smc/smc_ib.c | 1 +
net/smc/smc_pnet.c | 4 +++-
net/smc/smc_rx.c | 2 ++
net/smc/smc_tx.c | 18 ++++++++++--------
net/smc/smc_wr.c | 2 +-
13 files changed, 66 insertions(+), 45 deletions(-)
--
2.13.5
--
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: Latest net-next from GIT panic
From: Paweł Staszewski @ 2017-09-20 11:02 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Linux Kernel Network Developers
In-Reply-To: <7fee43ee-75b5-c2f5-cf8d-684ceefcd2d1@itcare.pl>
Ok resumed and soo far:
Panic:
# bad: [9cc9a5cb176ccb4f2cda5ac34da5a659926f125f] datapath: Avoid using
stack larger than 1024.
git bisect bad 9cc9a5cb176ccb4f2cda5ac34da5a659926f125f
No panic:
# good: [073cf9e20c333ab29744717a23f9e43ec7512a20] Merge branch
'udp-reduce-cache-pressure'
git bisect good 073cf9e20c333ab29744717a23f9e43ec7512a20
W dniu 2017-09-20 o 12:22, Paweł Staszewski pisze:
> Soo far bisected and marked:
>
> git bisect start
> # bad: [07dd6cc1fff160143e82cf5df78c1db0b6e03355] Linux 4.13.2
> git bisect bad 07dd6cc1fff160143e82cf5df78c1db0b6e03355
> # good: [5d7d2e03e0f01a992e3521b180c3d3e67905f269] Linux 4.12.13
> git bisect good 5d7d2e03e0f01a992e3521b180c3d3e67905f269
> # good: [6f7da290413ba713f0cdd9ff1a2a9bb129ef4f6c] Linux 4.12
> git bisect good 6f7da290413ba713f0cdd9ff1a2a9bb129ef4f6c
> # bad: [ac7b75966c9c86426b55fe1c50ae148aa4571075] Merge tag
> 'pinctrl-v4.13-1' of
> git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl
> git bisect bad ac7b75966c9c86426b55fe1c50ae148aa4571075
> # good: [e24dd9ee5399747b71c1d982a484fc7601795f31] Merge branch 'next'
> of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security
> git bisect good e24dd9ee5399747b71c1d982a484fc7601795f31
> # good: [e24dd9ee5399747b71c1d982a484fc7601795f31] Merge branch 'next'
> of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security
> git bisect good e24dd9ee5399747b71c1d982a484fc7601795f31
> # good: [e24dd9ee5399747b71c1d982a484fc7601795f31] Merge branch 'next'
> of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security
> git bisect good e24dd9ee5399747b71c1d982a484fc7601795f31
>
>
>
> W dniu 2017-09-20 o 12:21, Paweł Staszewski pisze:
>> Ok kernel crashed with different panic that i didnt catch when i was
>> doing bisect and now my bisection is broken :)
>>
>> git bisect good
>> Bisecting: 1787 revisions left to test after this (roughly 11 steps)
>> error: Your local changes to the following files would be overwritten
>> by checkout:
>> Documentation/00-INDEX
>> Documentation/ABI/stable/sysfs-class-udc
>> Documentation/ABI/testing/configfs-usb-gadget-uac1
>> Documentation/ABI/testing/ima_policy
>> Documentation/ABI/testing/sysfs-bus-iio
>> Documentation/ABI/testing/sysfs-bus-iio-meas-spec
>> Documentation/ABI/testing/sysfs-bus-iio-timer-stm32
>> Documentation/ABI/testing/sysfs-class-net
>> Documentation/ABI/testing/sysfs-class-power-twl4030
>> Documentation/ABI/testing/sysfs-class-typec
>> Documentation/DMA-API.txt
>> Documentation/IRQ-domain.txt
>> Documentation/Makefile
>> Documentation/PCI/MSI-HOWTO.txt
>> Documentation/RCU/00-INDEX
>> Documentation/RCU/Design/Requirements/Requirements.html
>> Documentation/RCU/checklist.txt
>> Documentation/admin-guide/README.rst
>> Documentation/admin-guide/devices.txt
>> Documentation/admin-guide/index.rst
>> Documentation/admin-guide/kernel-parameters.txt
>> Documentation/admin-guide/pm/cpufreq.rst
>> Documentation/admin-guide/pm/intel_pstate.rst
>> Documentation/admin-guide/ras.rst
>> Documentation/arm/Atmel/README
>> Documentation/block/biodoc.txt
>> Documentation/conf.py
>> Documentation/core-api/assoc_array.rst
>> Documentation/core-api/atomic_ops.rst
>> Documentation/core-api/index.rst
>> Documentation/crypto/asymmetric-keys.txt
>> Documentation/dev-tools/index.rst
>> Documentation/dev-tools/sparse.rst
>> Documentation/devicetree/bindings/arm/amlogic.txt
>> Documentation/devicetree/bindings/arm/atmel-at91.txt
>> Documentation/devicetree/bindings/arm/ccn.txt
>> Documentation/devicetree/bindings/arm/cpus.txt
>> Documentation/devicetree/bindings/arm/gemini.txt
>> Documentation/devicetree/bindings/arm/hisilicon/hisilicon.txt
>> Documentation/devicetree/bindings/arm/keystone/keystone.txt
>> Documentation/devicetree/bindings/arm/mediatek.txt
>> Documentation/devicetree/bindings/arm/rockchip.txt
>> Documentation/devicetree/bindings/arm/shmobile.txt
>> Documentation/devicetree/bindings/arm/tegra.txt
>> Documentation/devicetree/bindings/ata/ahci-fsl-qoriq.txt
>> Documentation/devicetree/bindings/bus/brcm,gisb-arb.txt
>> Documentation/devicetree/bindings/clock/brcm,iproc-clocks.txt
>> Documentation/devicetree/bindings/cpufreq/ti-cpufreq.txt
>> Documentation/devicetree/bindings/gpio/gpio_atmel.txt
>> Documentation/devicetree/bindings/iio/adc/amlogic,meson-saradc.txt
>> Documentation/devicetree/bindings/iio/adc/renesas,gyroadc.txt
>> Documentation/devicetree/bindings/iio/adc/st,stm32-adc.txt
>> Documentation/devicetree/bindings/iio/imu/st_lsm6dsx.txt
>> Documentation/devicetree/bindings/interrupt-controller/allwinner,sunxi-nmi.txt
>>
>> Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2400-vic.txt
>>
>> Documentation/devicetree/bindings/interrupt-controller/mediatek,sysirq.txt
>>
>> Documentation/devicetree/bindings/leds/common.txt
>> Documentation/devicetree/bindings/mfd/hi6421.txt
>> Documentation/devicetree/bindings/mfd/tps65910.txt
>> Documentation/devicetree/bindings/mmc/fsl-esdhc.txt
>> Documentation/devicetree/bindings/mmc/k3-dw-mshc.txt
>> Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt
>> Documentation/devicetree/bindings/mmc/ti-omap-hsmmc.txt
>> Documentation/devicetree/bindings/mtd/atmel-nand.txt
>> Documentation/devicetree/bindings/net/dsa/b53.txt
>> Documentation/devicetree/bindings/net/ethernet.txt
>> Documentation/devicetree/bindings/net/macb.txt
>> Documentation/devicetree/bindings/net/marvell-orion-mdio.txt
>> Documentation/devicetree/bindings/net/ti,wilink-st.txt
>> Documentation/devicetree/bindings/net/wireless/ti,wlcore.txt
>> Documentation/devicetree/bindings/nvmem/rockchip-efuse.txt
>> Documentation/devicetree/bindings/opp/opp.txt
>> Documentation/devicetree/bindings/phy/bcm-ns-usb3-phy.txt
>> Documentation/devicetree/bindings/phy/brcm-sata-phy.txt
>> Documentation/devicetree/bindings/phy/meson8b-usb2-phy.txt
>> Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
>> Documentation/devicetree/bindings/power/rockchip-io-domain.txt
>> Documentation/devicetree/bindings/power/supply/bq27xxx.txt
>> Documentation/devicetree/bindings/property-units.txt
>> Documentation/devicetree/bindings/regulator/regulator.txt
>> Documentation/devicetree/bindings/serial/8
>> error: The following untracked working tree files would be
>> overwritten by checkout:
>> Documentation/ABI/testing/sysfs-class-net-phydev
>> Documentation/DocBook/.gitignore
>> Documentation/DocBook/Makefile
>> Documentation/DocBook/filesystems.tmpl
>> Documentation/DocBook/kernel-hacking.tmpl
>> Documentation/DocBook/kernel-locking.tmpl
>> Documentation/DocBook/kgdb.tmpl
>> Documentation/DocBook/libata.tmpl
>> Documentation/DocBook/librs.tmpl
>> Documentation/DocBook/lsm.tmpl
>> Documentation/DocBook/mtdnand.tmpl
>> Documentation/DocBook/networking.tmpl
>> Documentation/DocBook/rapidio.tmpl
>> Documentation/DocBook/s390-drivers.tmpl
>> Documentation/DocBook/scsi.tmpl
>> Documentation/DocBook/sh.tmpl
>> Documentation/DocBook/stylesheet.xsl
>> Documentation/DocBook/w1.tmpl
>> Documentation/DocBook/z8530book.tmpl
>> Documentation/Makefile.sphinx
>> Documentation/RCU/trace.txt
>> Documentation/devicetree/bindings/i2c/i2c-mt6577.txt
>> Documentation/devicetree/bindings/misc/allwinner,syscon.txt
>> Documentation/devicetree/bindings/net/cortina.txt
>> Documentation/devicetree/bindings/net/dsa/ksz.txt
>> Documentation/devicetree/bindings/net/dwmac-sun8i.txt
>> Documentation/devicetree/bindings/net/qca,qca7000.txt
>> Documentation/devicetree/bindings/power/max8903-charger.txt
>> Documentation/devicetree/bindings/power_supply/maxim,max14656.txt
>> Documentation/devicetree/bindings/ptp/brcm,ptp-dte.txt
>> Documentation/devicetree/bindings/timer/moxa,moxart-timer.txt
>> Documentation/doc-guide/docbook.rst
>> Documentation/networking/tls.txt
>> Documentation/prctl/no_new_privs.txt
>> Documentation/prctl/seccomp_filter.txt
>> Documentation/security/00-INDEX
>> Documentation/security/IMA-templates.txt
>> Documentation/security/LSM.txt
>> Documentation/security/LoadPin.txt
>> Documentation/security/SELinux.txt
>> Documentation/security/Smack.txt
>> Documentation/security/Yama.txt
>> Documentation/security/apparmor.txt
>> Documentation/security/conf.py
>> Documentation/security/credentials.txt
>> Documentation/security/keys-ecryptfs.txt
>> Documentation/security/keys-request-key.txt
>> Documentation/security/keys-trusted-encrypted.txt
>> Documentation/security/keys.txt
>> Documentation/security/self-protection.txt
>> Documentation/security/tomoyo.txt
>> Documentation/sphinx/convert_template.sed
>> Documentation/sphinx/post_convert.sed
>> Documentation/sphinx/tmplcvt
>> Documentation/usb/typec.rst
>> Documentation/usb/usb3-debug-port.rst
>> arch/arm/boot/dts/rk1108-evb.dts
>> arch/arm/boot/dts/rk1108.dtsi
>> arch/arm/boot/dts/tegra20-whistler.dts
>> arch/arm/mach-omap2/opp.c
>> arch/arm/mach-omap2/pmu.c
>> arch/ia64/include/asm/siginfo.h
>> arch/m32r/include/uapi/asm/siginfo.h
>> arch/microblaze/include/asm/bitops.h
>> arch/microblaze/include/asm/bug.h
>> arch/microblaze/include/asm/bugs.h
>> arch/microblaze/include/asm/div64.h
>> arch/microblaze/include/asm/emergency-restart.h
>> arch/microblaze/include/asm/fb.h
>> arch/microblaze/include/asm/hardirq.h
>> arch/microblaze/include/asm/irq_regs.h
>> arch/microblaze/include/asm/kdebug.h
>> arch/microblaze/include/asm/kmap_types.h
>> arch/microblaze/include/asm/linkage.h
>> arch/microblaze/include/asm/local.h
>> arch/microblaze/include/asm/local64.h
>> arch/microblaze/include/asm/parport.h
>> arch/microblaze/include/asm/percpu.h
>> arch/microblaze/include/asm/serial.h
>> arch/microblaze/include/asm/shmparam.h
>> arch/microblaze/include/asm/topology.h
>> arch/microblaze/include/asm/ucontext.h
>> arch/microblaze/include/asm/vga.h
>> arch/microblaze/include/asm/xor.h
>> arch/microblaze/include/uapi/asm/bitsperlong.h
>> arch/microblaze/include/uapi/asm/errno.h
>> arch/microblaze/include/uapi/asm/fcntl.h
>> arch/microblaze/include/uapi/asm/ioctl.h
>> arch/microblaze/include/uapi/asm/ioctls.h
>> arch/microblaze/include/uapi/asm/ipcbuf.h
>> arch/microblaze/include/uapi/asm/kvm_para.h
>> arch/microblaze/include/uapi/asm/mman.h
>> arch/microblaze/include/uapi/asm/msgbuf.h
>> arch/microblaze/include/uapi/asm/param.h
>> arch/microblaze/include/uapi/asm/poll.h
>> arch/microblaze/include/uapi/asm/resource.h
>> arch/microblaze/include/uapi/asm/sembuf.h
>> arch/microblaze/include/uapi/asm/shmbuf.h
>> arch/microblaze/include/uapi/asm/siginfo.h
>> arch/microblaze/include/uapi/asm/signal.h
>> arch/microblaze/includ
>> Aborting
>>
>>
>>
>> W dniu 2017-09-20 o 11:45, Paweł Staszewski pisze:
>>> Ok looks like ending bisection
>>>
>>>
>>> Latest bisected kernel when there is no kernel panic 4.12.0+ (from
>>> next) - but only this warning:
>>>
>>> [ 309.030019] NETDEV WATCHDOG: enp4s0f0 (ixgbe): transmit queue 0
>>> timed out
>>> [ 309.030034] ------------[ cut here ]------------
>>> [ 309.030040] WARNING: CPU: 35 PID: 0 at dev_watchdog+0xcf/0x139
>>> [ 309.030041] Modules linked in: bonding ipmi_si x86_pkg_temp_thermal
>>> [ 309.030045] CPU: 35 PID: 0 Comm: swapper/35 Not tainted 4.12.0+ #5
>>> [ 309.030046] task: ffff88086d98a000 task.stack: ffffc90003378000
>>> [ 309.030048] RIP: 0010:dev_watchdog+0xcf/0x139
>>> [ 309.030049] RSP: 0018:ffff88087fbc3ea8 EFLAGS: 00010246
>>> [ 309.030050] RAX: 000000000000003d RBX: ffff88046b680000 RCX:
>>> 0000000000000000
>>> [ 309.030050] RDX: ffff88087fbd2f01 RSI: 0000000000000000 RDI:
>>> ffff88087fbcda08
>>> [ 309.030051] RBP: ffff88087fbc3eb8 R08: 0000000000000000 R09:
>>> ffff88087ff80a04
>>> [ 309.030051] R10: 0000000000000000 R11: ffff88086d98a001 R12:
>>> 0000000000000000
>>> [ 309.030052] R13: ffff88087fbc3ef8 R14: ffff88086d98a000 R15:
>>> ffffffff81c06008
>>> [ 309.030053] FS: 0000000000000000(0000) GS:ffff88087fbc0000(0000)
>>> knlGS:0000000000000000
>>> [ 309.030054] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> [ 309.030054] CR2: 00007fba600f6098 CR3: 000000086b955000 CR4:
>>> 00000000001406e0
>>> [ 309.030055] Call Trace:
>>> [ 309.030057] <IRQ>
>>> [ 309.030059] ? netif_tx_lock+0x79/0x79
>>> [ 309.030062] call_timer_fn.isra.24+0x17/0x77
>>> [ 309.030063] run_timer_softirq+0x118/0x161
>>> [ 309.030065] ? netif_tx_lock+0x79/0x79
>>> [ 309.030066] ? ktime_get+0x2b/0x42
>>> [ 309.030070] ? lapic_next_deadline+0x21/0x27
>>> [ 309.030073] ? clockevents_program_event+0xa8/0xc5
>>> [ 309.030076] __do_softirq+0xa8/0x19d
>>> [ 309.030078] irq_exit+0x5d/0x6b
>>> [ 309.030079] smp_apic_timer_interrupt+0x2a/0x36
>>> [ 309.030082] apic_timer_interrupt+0x89/0x90
>>> [ 309.030085] RIP: 0010:mwait_idle+0x4e/0x6a
>>> [ 309.030086] RSP: 0018:ffffc9000337be98 EFLAGS: 00000246 ORIG_RAX:
>>> ffffffffffffff10
>>> [ 309.030087] RAX: 0000000000000000 RBX: 0000000000000000 RCX:
>>> 0000000000000000
>>> [ 309.030087] RDX: 0000000000000000 RSI: 0000000000000000 RDI:
>>> ffff88086d98a000
>>> [ 309.030088] RBP: ffffc9000337be98 R08: ffff88046f8279a0 R09:
>>> ffff88046f827040
>>> [ 309.030089] R10: ffff88086d98a000 R11: ffff88086d98a000 R12:
>>> 0000000000000000
>>> [ 309.030089] R13: ffff88086d98a000 R14: ffff88086d98a000 R15:
>>> ffff88086d98a000
>>> [ 309.030090] </IRQ>
>>> [ 309.030094] arch_cpu_idle+0xa/0xc
>>> [ 309.030095] default_idle_call+0x19/0x1b
>>> [ 309.030102] do_idle+0xbc/0x196
>>> [ 309.030104] cpu_startup_entry+0x1d/0x20
>>> [ 309.030105] start_secondary+0xd8/0xdc
>>> [ 309.030108] secondary_startup_64+0x9f/0x9f
>>> [ 309.030109] Code: cc 75 bd eb 35 48 89 df c6 05 c3 dc 74 00 01 e8
>>> 3a 62 fe ff 44 89 e1 48 89 de 48 89 c2 48 c7 c7 0f 65 a4 81 31 c0 e8
>>> 3d 4c b5 ff <0f> ff 48 8b 83 e0 01 00 00 48 89 df ff 50 78 48 8b 05
>>> a0 bc 6a
>>> [ 309.030128] ---[ end trace 9102cb25703ae2d9 ]---
>>>
>>>
>>> I just marked it as good - cause this problem above is differend -
>>> and im going to:
>>>
>>> git bisect good
>>> Bisecting: 1787 revisions left to test after this (roughly 11 steps)
>>>
>>>
>>>
>>>
>>> W dniu 2017-09-20 o 10:44, Paweł Staszewski pisze:
>>>> Trying to make video from ipmi :)
>>>>
>>>> with that results:
>>>>
>>>> https://bugzilla.kernel.org/attachment.cgi?id=258521
>>>>
>>>> catched two more lines where it starts - panic from 4.13.2.
>>>>
>>>>
>>>> Now will try tro do some bisection
>>>>
>>>>
>>>>
>>>> W dniu 2017-09-20 o 09:58, Paweł Staszewski pisze:
>>>>> Hi
>>>>>
>>>>>
>>>>> Will try bisecting tonight
>>>>>
>>>>>
>>>>>
>>>>> W dniu 2017-09-20 o 05:24, Eric Dumazet pisze:
>>>>>> On Wed, 2017-09-20 at 02:06 +0200, Paweł Staszewski wrote:
>>>>>>> Just checked kernel 4.13.2 and same problem
>>>>>>>
>>>>>>> Just after start all 6 bgp sessions - and kernel starts to learn
>>>>>>> routes
>>>>>>> it panic.
>>>>>>>
>>>>>>> https://bugzilla.kernel.org/attachment.cgi?id=258509
>>>>>>>
>>>>>>
>>>>>> Unfortunately we have not enough information from these traces.
>>>>>>
>>>>>> Can you get a full stack trace ?
>>>>>>
>>>>>> Alternatively, can you bisect ?
>>>>>>
>>>>>> Thanks.
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
^ permalink raw reply
* [PATCH net 9/9] net: hns3: Fix for pri to tc mapping in TM
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
Current mapping between pri and tc is one to one,
so user can't map multi priorities to the same tc.
This patch changes the mapping to many to one.
Fixes: 848440544b41f ("net: hns3: Add support of TX Scheduler & Shaper to HNS3 driver")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 3 ++-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 2 +-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 16 +++++++++-------
3 files changed, 12 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index ad685f5..1a01cad 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -376,12 +376,12 @@ struct hnae3_ae_algo {
struct hnae3_tc_info {
u16 tqp_offset; /* TQP offset from base TQP */
u16 tqp_count; /* Total TQPs */
- u8 up; /* user priority */
u8 tc; /* TC index */
bool enable; /* If this TC is enable or not */
};
#define HNAE3_MAX_TC 8
+#define HNAE3_MAX_USER_PRIO 8
struct hnae3_knic_private_info {
struct net_device *netdev; /* Set by KNIC client when init instance */
u16 rss_size; /* Allocated RSS queues */
@@ -389,6 +389,7 @@ struct hnae3_knic_private_info {
u16 num_desc;
u8 num_tc; /* Total number of enabled TCs */
+ u8 prio_tc[HNAE3_MAX_USER_PRIO]; /* TC indexed by prio */
struct hnae3_tc_info tc_info[HNAE3_MAX_TC]; /* Idx of array is HW TC */
u16 num_tqps; /* total number of TQPs in this handle */
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index 7f8dd12..9fcfd93 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -176,7 +176,6 @@ struct hclge_pg_info {
struct hclge_tc_info {
u8 tc_id;
u8 tc_sch_mode; /* 0: sp; 1: dwrr */
- u8 up;
u8 pgid;
u32 bw_limit;
};
@@ -197,6 +196,7 @@ struct hclge_tm_info {
u8 num_tc;
u8 num_pg; /* It must be 1 if vNET-Base schd */
u8 pg_dwrr[HCLGE_PG_NUM];
+ u8 prio_tc[HNAE3_MAX_USER_PRIO];
struct hclge_pg_info pg_info[HCLGE_PG_NUM];
struct hclge_tc_info tc_info[HNAE3_MAX_TC];
enum hclge_fc_mode fc_mode;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index b7ba7aa..73a75d7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -128,9 +128,7 @@ static int hclge_fill_pri_array(struct hclge_dev *hdev, u8 *pri, u8 pri_id)
{
u8 tc;
- for (tc = 0; tc < hdev->tm_info.num_tc; tc++)
- if (hdev->tm_info.tc_info[tc].up == pri_id)
- break;
+ tc = hdev->tm_info.prio_tc[pri_id];
if (tc >= hdev->tm_info.num_tc)
return -EINVAL;
@@ -158,7 +156,7 @@ static int hclge_up_to_tc_map(struct hclge_dev *hdev)
hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_PRI_TO_TC_MAPPING, false);
- for (pri_id = 0; pri_id < hdev->tm_info.num_tc; pri_id++) {
+ for (pri_id = 0; pri_id < HNAE3_MAX_USER_PRIO; pri_id++) {
ret = hclge_fill_pri_array(hdev, pri, pri_id);
if (ret)
return ret;
@@ -405,16 +403,17 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport)
kinfo->tc_info[i].tqp_offset = i * kinfo->rss_size;
kinfo->tc_info[i].tqp_count = kinfo->rss_size;
kinfo->tc_info[i].tc = i;
- kinfo->tc_info[i].up = hdev->tm_info.tc_info[i].up;
} else {
/* Set to default queue if TC is disable */
kinfo->tc_info[i].enable = false;
kinfo->tc_info[i].tqp_offset = 0;
kinfo->tc_info[i].tqp_count = 1;
kinfo->tc_info[i].tc = 0;
- kinfo->tc_info[i].up = 0;
}
}
+
+ memcpy(kinfo->prio_tc, hdev->tm_info.prio_tc,
+ FIELD_SIZEOF(struct hnae3_knic_private_info, prio_tc));
}
static void hclge_tm_vport_info_update(struct hclge_dev *hdev)
@@ -436,12 +435,15 @@ static void hclge_tm_tc_info_init(struct hclge_dev *hdev)
for (i = 0; i < hdev->tm_info.num_tc; i++) {
hdev->tm_info.tc_info[i].tc_id = i;
hdev->tm_info.tc_info[i].tc_sch_mode = HCLGE_SCH_MODE_DWRR;
- hdev->tm_info.tc_info[i].up = i;
hdev->tm_info.tc_info[i].pgid = 0;
hdev->tm_info.tc_info[i].bw_limit =
hdev->tm_info.pg_info[0].bw_limit;
}
+ for (i = 0; i < HNAE3_MAX_USER_PRIO; i++)
+ hdev->tm_info.prio_tc[i] =
+ (i >= hdev->tm_info.num_tc) ? 0 : i;
+
hdev->flag &= ~HCLGE_FLAG_DCB_ENABLE;
}
--
1.9.1
^ permalink raw reply related
* [PATCH net 8/9] net: hns3: Fix for setting rss_size incorrectly
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
rss_size is 1, 2, 4, 8, 16, 32, 64, 128, but acutal tc queue
size can be any u16 less than 128. If tc queue size is 5, we
set the rss_size to 8, indirection table will be used to limit
the size of actual queue size.
It may cause dropping of receiving packet in hardware if
rss_size is not set correctly.
For now, each TC has the same rss size.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 76 ++++++++++------------
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 1 +
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 1 +
3 files changed, 38 insertions(+), 40 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 1e15ce1..d27618b 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -2606,6 +2606,7 @@ static int hclge_rss_init_hw(struct hclge_dev *hdev)
u16 tc_valid[HCLGE_MAX_TC_NUM];
u16 tc_size[HCLGE_MAX_TC_NUM];
u32 *rss_indir = NULL;
+ u16 rss_size = 0, roundup_size;
const u8 *key;
int i, ret, j;
@@ -2620,7 +2621,13 @@ static int hclge_rss_init_hw(struct hclge_dev *hdev)
for (j = 0; j < hdev->num_vmdq_vport + 1; j++) {
for (i = 0; i < HCLGE_RSS_IND_TBL_SIZE; i++) {
vport[j].rss_indirection_tbl[i] =
- i % hdev->rss_size_max;
+ i % vport[j].alloc_rss_size;
+
+ /* vport 0 is for PF */
+ if (j != 0)
+ continue;
+
+ rss_size = vport[j].alloc_rss_size;
rss_indir[i] = vport[j].rss_indirection_tbl[i];
}
}
@@ -2637,42 +2644,31 @@ static int hclge_rss_init_hw(struct hclge_dev *hdev)
if (ret)
goto err;
+ /* Each TC have the same queue size, and tc_size set to hardware is
+ * the log2 of roundup power of two of rss_size, the acutal queue
+ * size is limited by indirection table.
+ */
+ if (rss_size > HCLGE_RSS_TC_SIZE_7 || rss_size == 0) {
+ dev_err(&hdev->pdev->dev,
+ "Configure rss tc size failed, invalid TC_SIZE = %d\n",
+ rss_size);
+ return -EINVAL;
+ }
+
+ roundup_size = roundup_pow_of_two(rss_size);
+ roundup_size = ilog2(roundup_size);
+
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
- if (hdev->hw_tc_map & BIT(i))
- tc_valid[i] = 1;
- else
- tc_valid[i] = 0;
+ tc_valid[i] = 0;
- switch (hdev->rss_size_max) {
- case HCLGE_RSS_TC_SIZE_0:
- tc_size[i] = 0;
- break;
- case HCLGE_RSS_TC_SIZE_1:
- tc_size[i] = 1;
- break;
- case HCLGE_RSS_TC_SIZE_2:
- tc_size[i] = 2;
- break;
- case HCLGE_RSS_TC_SIZE_3:
- tc_size[i] = 3;
- break;
- case HCLGE_RSS_TC_SIZE_4:
- tc_size[i] = 4;
- break;
- case HCLGE_RSS_TC_SIZE_5:
- tc_size[i] = 5;
- break;
- case HCLGE_RSS_TC_SIZE_6:
- tc_size[i] = 6;
- break;
- case HCLGE_RSS_TC_SIZE_7:
- tc_size[i] = 7;
- break;
- default:
- break;
- }
- tc_offset[i] = hdev->rss_size_max * i;
+ if (!(hdev->hw_tc_map & BIT(i)))
+ continue;
+
+ tc_valid[i] = 1;
+ tc_size[i] = roundup_size;
+ tc_offset[i] = rss_size * i;
}
+
ret = hclge_set_rss_tc_mode(hdev, tc_valid, tc_size, tc_offset);
err:
@@ -4166,12 +4162,6 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
return ret;
}
- ret = hclge_rss_init_hw(hdev);
- if (ret) {
- dev_err(&pdev->dev, "Rss init fail, ret =%d\n", ret);
- return ret;
- }
-
ret = hclge_init_vlan_config(hdev);
if (ret) {
dev_err(&pdev->dev, "VLAN init fail, ret =%d\n", ret);
@@ -4184,6 +4174,12 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
return ret;
}
+ ret = hclge_rss_init_hw(hdev);
+ if (ret) {
+ dev_err(&pdev->dev, "Rss init fail, ret =%d\n", ret);
+ return ret;
+ }
+
setup_timer(&hdev->service_timer, hclge_service_timer,
(unsigned long)hdev);
INIT_WORK(&hdev->service_task, hclge_service_task);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index edb10ad..7f8dd12 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -477,6 +477,7 @@ struct hclge_vport {
u8 rss_hash_key[HCLGE_RSS_KEY_SIZE]; /* User configured hash keys */
/* User configured lookup table entries */
u8 rss_indirection_tbl[HCLGE_RSS_IND_TBL_SIZE];
+ u16 alloc_rss_size;
u16 qs_offset;
u16 bw_limit; /* VSI BW Limit (0 = disabled) */
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index fe659f7..b7ba7aa 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -397,6 +397,7 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport)
kinfo->num_tqps / kinfo->num_tc);
vport->qs_offset = hdev->tm_info.num_tc * vport->vport_id;
vport->dwrr = 100; /* 100 percent as init */
+ vport->alloc_rss_size = kinfo->rss_size;
for (i = 0; i < kinfo->num_tc; i++) {
if (hdev->hw_tc_map & BIT(i)) {
--
1.9.1
^ permalink raw reply related
* [PATCH net 7/9] net: hns3: Fix typo error for feild in hclge_tm
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
This patch fixes a typo error for feild, which should be field.
Fixes: 848440544b41f ("net: hns3: Add support of TX Scheduler & Shaper to HNS3 driver")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 20 ++++++++++----------
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 4 ++--
2 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index c91dbf1..fe659f7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -280,11 +280,11 @@ static int hclge_tm_pg_shapping_cfg(struct hclge_dev *hdev,
shap_cfg_cmd->pg_id = pg_id;
- hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_B, ir_b);
- hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_U, ir_u);
- hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_S, ir_s);
- hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, BS_B, bs_b);
- hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, BS_S, bs_s);
+ hclge_tm_set_field(shap_cfg_cmd->pg_shapping_para, IR_B, ir_b);
+ hclge_tm_set_field(shap_cfg_cmd->pg_shapping_para, IR_U, ir_u);
+ hclge_tm_set_field(shap_cfg_cmd->pg_shapping_para, IR_S, ir_s);
+ hclge_tm_set_field(shap_cfg_cmd->pg_shapping_para, BS_B, bs_b);
+ hclge_tm_set_field(shap_cfg_cmd->pg_shapping_para, BS_S, bs_s);
return hclge_cmd_send(&hdev->hw, &desc, 1);
}
@@ -307,11 +307,11 @@ static int hclge_tm_pri_shapping_cfg(struct hclge_dev *hdev,
shap_cfg_cmd->pri_id = pri_id;
- hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_B, ir_b);
- hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_U, ir_u);
- hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_S, ir_s);
- hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, BS_B, bs_b);
- hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, BS_S, bs_s);
+ hclge_tm_set_field(shap_cfg_cmd->pri_shapping_para, IR_B, ir_b);
+ hclge_tm_set_field(shap_cfg_cmd->pri_shapping_para, IR_U, ir_u);
+ hclge_tm_set_field(shap_cfg_cmd->pri_shapping_para, IR_S, ir_s);
+ hclge_tm_set_field(shap_cfg_cmd->pri_shapping_para, BS_B, bs_b);
+ hclge_tm_set_field(shap_cfg_cmd->pri_shapping_para, BS_S, bs_s);
return hclge_cmd_send(&hdev->hw, &desc, 1);
}
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
index 7e67337..85158b0 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h
@@ -94,10 +94,10 @@ struct hclge_bp_to_qs_map_cmd {
u32 rsvd1;
};
-#define hclge_tm_set_feild(dest, string, val) \
+#define hclge_tm_set_field(dest, string, val) \
hnae_set_field((dest), (HCLGE_TM_SHAP_##string##_MSK), \
(HCLGE_TM_SHAP_##string##_LSH), val)
-#define hclge_tm_get_feild(src, string) \
+#define hclge_tm_get_field(src, string) \
hnae_get_field((src), (HCLGE_TM_SHAP_##string##_MSK), \
(HCLGE_TM_SHAP_##string##_LSH))
--
1.9.1
^ permalink raw reply related
* [PATCH net 6/9] net: hns3: Fix for rx priv buf allocation when DCB is not supported
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
When hdev doesn't support DCB, rx private buffer is not allocated,
otherwise there is not enough buffer for rx shared buffer, causing
buffer allocation process to fail.
This patch fixes by checking the dcb capability in
hclge_rx_buffer_calc.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index c08c41f..1e15ce1 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1489,6 +1489,16 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev, u32 tx_size)
struct hclge_priv_buf *priv;
int i;
+ /* When DCB is not supported, rx private
+ * buffer is not allocated.
+ */
+ if (!hnae3_dev_dcb_supported(hdev)) {
+ if (!hclge_is_rx_buf_ok(hdev, rx_all))
+ return -ENOMEM;
+
+ return 0;
+ }
+
/* step 1, try to alloc private buffer for all enabled tc */
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
priv = &hdev->priv_buf[i];
--
1.9.1
^ permalink raw reply related
* [PATCH net 5/9] net: hns3: Fix for rx_priv_buf_alloc not setting rx shared buffer
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
rx_priv_buf_alloc is used to tell hardware how much buffer is
used for rx direction, right now only the private buffer is
assigned.
For ae_dev that doesn't support DCB, private rx buffer is assigned
to zero, only shared rx buffer is used. So not setting the shared
rx buffer cause dropping of packet in SSU.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 3 ++-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 4 ++++
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index 30e2ad5..758cf39 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -270,7 +270,8 @@ struct hclge_tx_buff_alloc {
struct hclge_rx_priv_buff {
__le16 buf_num[HCLGE_TC_NUM];
- u8 rsv[8];
+ __le16 shared_buf;
+ u8 rsv[6];
};
struct hclge_query_version {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index bf3179a..c08c41f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1622,6 +1622,10 @@ static int hclge_rx_priv_buf_alloc(struct hclge_dev *hdev)
cpu_to_le16(true << HCLGE_TC0_PRI_BUF_EN_B);
}
+ req->shared_buf =
+ cpu_to_le16((hdev->s_buf.buf_size >> HCLGE_BUF_UNIT_S) |
+ (1 << HCLGE_TC0_PRI_BUF_EN_B));
+
ret = hclge_cmd_send(&hdev->hw, &desc, 1);
if (ret) {
dev_err(&hdev->pdev->dev,
--
1.9.1
^ permalink raw reply related
* [PATCH net 4/9] net: hns3: Fix for not setting rx private buffer size to zero
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
When rx private buffer is disabled, there may be some case that
the rx private buffer is not set to zero, which may cause buffer
allocation process to fail.
This patch fixes this problem by setting priv->enable to 0 and
priv->buf_size to zero when rx private buffer is disabled.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 16 ++++++++++++++--
1 file changed, 14 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 1876418..bf3179a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1504,6 +1504,11 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev, u32 tx_size)
priv->wl.high = 2 * hdev->mps;
priv->buf_size = priv->wl.high;
}
+ } else {
+ priv->enable = 0;
+ priv->wl.low = 0;
+ priv->wl.high = 0;
+ priv->buf_size = 0;
}
}
@@ -1516,8 +1521,15 @@ int hclge_rx_buffer_calc(struct hclge_dev *hdev, u32 tx_size)
for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
priv = &hdev->priv_buf[i];
- if (hdev->hw_tc_map & BIT(i))
- priv->enable = 1;
+ priv->enable = 0;
+ priv->wl.low = 0;
+ priv->wl.high = 0;
+ priv->buf_size = 0;
+
+ if (!(hdev->hw_tc_map & BIT(i)))
+ continue;
+
+ priv->enable = 1;
if (hdev->tm_info.hw_pfc_map & BIT(i)) {
priv->wl.low = 128;
--
1.9.1
^ permalink raw reply related
* [PATCH net 3/9] net: hns3: Fix for DEFAULT_DV when dev doesn't support DCB
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
When ae_dev doesn't support DCB, DEFAULT_DV must be set to
a lower value, otherwise the buffer allocation process will
fail.
This patch fix it by setting it to 30K bytes.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 1 +
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 6 +++++-
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index c2b613b..30e2ad5 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -688,6 +688,7 @@ struct hclge_reset_tqp_queue {
#define HCLGE_DEFAULT_TX_BUF 0x4000 /* 16k bytes */
#define HCLGE_TOTAL_PKT_BUF 0x108000 /* 1.03125M bytes */
#define HCLGE_DEFAULT_DV 0xA000 /* 40k byte */
+#define HCLGE_DEFAULT_NON_DCB_DV 0x7800 /* 30K byte */
#define HCLGE_TYPE_CRQ 0
#define HCLGE_TYPE_CSQ 1
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index c515b84..1876418 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1444,7 +1444,11 @@ static bool hclge_is_rx_buf_ok(struct hclge_dev *hdev, u32 rx_all)
tc_num = hclge_get_tc_num(hdev);
pfc_enable_num = hclge_get_pfc_enalbe_num(hdev);
- shared_buf_min = 2 * hdev->mps + HCLGE_DEFAULT_DV;
+ if (hnae3_dev_dcb_supported(hdev))
+ shared_buf_min = 2 * hdev->mps + HCLGE_DEFAULT_DV;
+ else
+ shared_buf_min = 2 * hdev->mps + HCLGE_DEFAULT_NON_DCB_DV;
+
shared_buf_tc = pfc_enable_num * hdev->mps +
(tc_num - pfc_enable_num) * hdev->mps / 2 +
hdev->mps;
--
1.9.1
^ permalink raw reply related
* [PATCH net 2/9] net: hns3: Fix initialization when cmd is not supported
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
When ae_dev doesn't support DCB, rx_priv_wl_config,
common_thrd_config and tm_qs_bp_cfg can't be called, otherwise
cmd return fail, which causes the hclge module initialization
process to fail.
This patch fix it by adding a DCB capability flag to check if
the ae_dev support DCB.
Fixes: 46a3df9f9718 ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 7 ++++++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 26 +++++++++++++---------
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 4 ++++
.../net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c | 10 ++++-----
4 files changed, 31 insertions(+), 16 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index 0f7b61a..ad685f5 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -50,10 +50,17 @@
#define HNAE3_DEV_INITED_B 0x0
#define HNAE3_DEV_SUPPORT_ROCE_B 0x1
+#define HNAE3_DEV_SUPPORT_DCB_B 0x2
+
+#define HNAE3_DEV_SUPPORT_ROCE_DCB_BITS (BIT(HNAE3_DEV_SUPPORT_DCB_B) |\
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B))
#define hnae3_dev_roce_supported(hdev) \
hnae_get_bit(hdev->ae_dev->flag, HNAE3_DEV_SUPPORT_ROCE_B)
+#define hnae3_dev_dcb_supported(hdev) \
+ hnae_get_bit(hdev->ae_dev->flag, HNAE3_DEV_SUPPORT_DCB_B)
+
#define ring_ptr_move_fw(ring, p) \
((ring)->p = ((ring)->p + 1) % (ring)->desc_num)
#define ring_ptr_move_bw(ring, p) \
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index eb78c23..c515b84 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1772,18 +1772,22 @@ int hclge_buffer_alloc(struct hclge_dev *hdev)
return ret;
}
- ret = hclge_rx_priv_wl_config(hdev);
- if (ret) {
- dev_err(&hdev->pdev->dev,
- "could not configure rx private waterline %d\n", ret);
- return ret;
- }
+ if (hnae3_dev_dcb_supported(hdev)) {
+ ret = hclge_rx_priv_wl_config(hdev);
+ if (ret) {
+ dev_err(&hdev->pdev->dev,
+ "could not configure rx private waterline %d\n",
+ ret);
+ return ret;
+ }
- ret = hclge_common_thrd_config(hdev);
- if (ret) {
- dev_err(&hdev->pdev->dev,
- "could not configure common threshold %d\n", ret);
- return ret;
+ ret = hclge_common_thrd_config(hdev);
+ if (ret) {
+ dev_err(&hdev->pdev->dev,
+ "could not configure common threshold %d\n",
+ ret);
+ return ret;
+ }
}
ret = hclge_common_wl_config(hdev);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
index 1c577d2..c91dbf1 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c
@@ -976,6 +976,10 @@ int hclge_pause_setup_hw(struct hclge_dev *hdev)
if (ret)
return ret;
+ /* Only DCB-supported dev supports qset back pressure setting */
+ if (!hnae3_dev_dcb_supported(hdev))
+ return 0;
+
for (i = 0; i < hdev->tm_info.num_tc; i++) {
ret = hclge_tm_qs_bp_cfg(hdev, i);
if (ret)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
index 94d8bb5..35369e1 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
@@ -42,15 +42,15 @@
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_GE), 0},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE), 0},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA),
- BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ HNAE3_DEV_SUPPORT_ROCE_DCB_BITS},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA_MACSEC),
- BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ HNAE3_DEV_SUPPORT_ROCE_DCB_BITS},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA),
- BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ HNAE3_DEV_SUPPORT_ROCE_DCB_BITS},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA_MACSEC),
- BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ HNAE3_DEV_SUPPORT_ROCE_DCB_BITS},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_MACSEC),
- BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ HNAE3_DEV_SUPPORT_ROCE_DCB_BITS},
/* required last entry */
{0, }
};
--
1.9.1
^ permalink raw reply related
* [PATCH net 1/9] net: hns3: Cleanup for ROCE capability flag in ae_dev
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
In-Reply-To: <1505904778-53217-1-git-send-email-linyunsheng@huawei.com>
This patch add the ROCE supported flag in the driver_data
field of pci_device_id, delete roce_pci_tbl and change
HNAE_DEV_SUPPORT_ROCE_B to HNAE3_DEV_SUPPORT_ROCE_B.
This cleanup is done in order to support adding capability
in pci_device_id and to fix initialization failure when
cmd is not supported.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 5 ++++-
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 25 ++++------------------
.../net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c | 16 +++++++++-----
3 files changed, 19 insertions(+), 27 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index b2f28ae..0f7b61a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -49,7 +49,10 @@
#define HNAE3_CLASS_NAME_SIZE 16
#define HNAE3_DEV_INITED_B 0x0
-#define HNAE_DEV_SUPPORT_ROCE_B 0x1
+#define HNAE3_DEV_SUPPORT_ROCE_B 0x1
+
+#define hnae3_dev_roce_supported(hdev) \
+ hnae_get_bit(hdev->ae_dev->flag, HNAE3_DEV_SUPPORT_ROCE_B)
#define ring_ptr_move_fw(ring, p) \
((ring)->p = ((ring)->p + 1) % (ring)->desc_num)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 44c722a..eb78c23 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -46,17 +46,7 @@ static int hclge_set_mta_filter_mode(struct hclge_dev *hdev,
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA), 0},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA_MACSEC), 0},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_MACSEC), 0},
- /* Required last entry */
- {0, }
-};
-
-static const struct pci_device_id roce_pci_tbl[] = {
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA_MACSEC), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA_MACSEC), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_MACSEC), 0},
- /* Required last entry */
+ /* required last entry */
{0, }
};
@@ -894,7 +884,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev)
hdev->num_tqps = __le16_to_cpu(req->tqp_num);
hdev->pkt_buf_size = __le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S;
- if (hnae_get_bit(hdev->ae_dev->flag, HNAE_DEV_SUPPORT_ROCE_B)) {
+ if (hnae3_dev_roce_supported(hdev)) {
hdev->num_roce_msix =
hnae_get_field(__le16_to_cpu(req->pf_intr_vector_number),
HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S);
@@ -3931,8 +3921,7 @@ static int hclge_init_client_instance(struct hnae3_client *client,
goto err;
if (hdev->roce_client &&
- hnae_get_bit(hdev->ae_dev->flag,
- HNAE_DEV_SUPPORT_ROCE_B)) {
+ hnae3_dev_roce_supported(hdev)) {
struct hnae3_client *rc = hdev->roce_client;
ret = hclge_init_roce_base_info(vport);
@@ -3955,8 +3944,7 @@ static int hclge_init_client_instance(struct hnae3_client *client,
break;
case HNAE3_CLIENT_ROCE:
- if (hnae_get_bit(hdev->ae_dev->flag,
- HNAE_DEV_SUPPORT_ROCE_B)) {
+ if (hnae3_dev_roce_supported(hdev)) {
hdev->roce_client = client;
vport->roce.client = client;
}
@@ -4068,7 +4056,6 @@ static void hclge_pci_uninit(struct hclge_dev *hdev)
static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
{
struct pci_dev *pdev = ae_dev->pdev;
- const struct pci_device_id *id;
struct hclge_dev *hdev;
int ret;
@@ -4083,10 +4070,6 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
hdev->ae_dev = ae_dev;
ae_dev->priv = hdev;
- id = pci_match_id(roce_pci_tbl, ae_dev->pdev);
- if (id)
- hnae_set_bit(ae_dev->flag, HNAE_DEV_SUPPORT_ROCE_B, 1);
-
ret = hclge_pci_init(hdev);
if (ret) {
dev_err(&pdev->dev, "PCI init failed\n");
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
index 4d68d6e..94d8bb5 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c
@@ -41,11 +41,16 @@
static const struct pci_device_id hns3_pci_tbl[] = {
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_GE), 0},
{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA_MACSEC), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA_MACSEC), 0},
- {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_MACSEC), 0},
+ {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA),
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE_RDMA_MACSEC),
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA),
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_50GE_RDMA_MACSEC),
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
+ {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_MACSEC),
+ BIT(HNAE3_DEV_SUPPORT_ROCE_B)},
/* required last entry */
{0, }
};
@@ -1348,6 +1353,7 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
}
ae_dev->pdev = pdev;
+ ae_dev->flag = ent->driver_data;
ae_dev->dev_type = HNAE3_DEV_KNIC;
pci_set_drvdata(pdev, ae_dev);
--
1.9.1
^ permalink raw reply related
* [PATCH net 0/9] TM related bugfixes for the HNS3 Ethernet Driver
From: Yunsheng Lin @ 2017-09-20 10:52 UTC (permalink / raw)
To: davem
Cc: huangdaode, xuwei5, liguozhu, Yisen.Zhuang, gabriele.paoloni,
john.garry, linuxarm, yisen.zhuang, salil.mehta, lipeng321,
netdev, linux-kernel
This patch set contains a few bugfixes related to hclge_tm module.
Yunsheng Lin (9):
net: hns3: Cleanup for ROCE capability flag in ae_dev
net: hns3: Fix initialization when cmd is not supported
net: hns3: Fix for DEFAULT_DV when dev doesn't support DCB
net: hns3: Fix for not setting rx private buffer size to zero
net: hns3: Fix for rx_priv_buf_alloc not setting rx shared buffer
net: hns3: Fix for rx priv buf allocation when DCB is not supported
net: hns3: Fix typo error for feild in hclge_tm
net: hns3: Fix for setting rss_size incorrectly
net: hns3: Fix for pri to tc mapping in TM
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 15 +-
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 4 +-
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 163 +++++++++++----------
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 3 +-
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 41 +++---
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 4 +-
.../net/ethernet/hisilicon/hns3/hns3pf/hns3_enet.c | 16 +-
7 files changed, 143 insertions(+), 103 deletions(-)
--
1.9.1
^ permalink raw reply
* [PATCH 1/2] netfilter: nat: Do not use ARRAY_SIZE() on spinlocks to fix zero div
From: Pablo Neira Ayuso @ 2017-09-20 10:49 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <1505904543-10004-1-git-send-email-pablo@netfilter.org>
From: Geert Uytterhoeven <geert@linux-m68k.org>
If no spinlock debugging options (CONFIG_GENERIC_LOCKBREAK,
CONFIG_DEBUG_SPINLOCK, CONFIG_DEBUG_LOCK_ALLOC) are enabled on a UP
platform (e.g. m68k defconfig), arch_spinlock_t is an empty struct,
hence using ARRAY_SIZE(nf_nat_locks) causes a division by zero:
net/netfilter/nf_nat_core.c: In function ‘nf_nat_setup_info’:
net/netfilter/nf_nat_core.c:432: warning: division by zero
net/netfilter/nf_nat_core.c: In function ‘__nf_nat_cleanup_conntrack’:
net/netfilter/nf_nat_core.c:535: warning: division by zero
net/netfilter/nf_nat_core.c:537: warning: division by zero
net/netfilter/nf_nat_core.c: In function ‘nf_nat_init’:
net/netfilter/nf_nat_core.c:810: warning: division by zero
net/netfilter/nf_nat_core.c:811: warning: division by zero
net/netfilter/nf_nat_core.c:824: warning: division by zero
Fix this by using the CONNTRACK_LOCKS definition instead.
Suggested-by: Florian Westphal <fw@strlen.de>
Fixes: 8073e960a03bf7b5 ("netfilter: nat: use keyed locks")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
net/netfilter/nf_nat_core.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
index f393a7086025..af8345fc4fbd 100644
--- a/net/netfilter/nf_nat_core.c
+++ b/net/netfilter/nf_nat_core.c
@@ -429,7 +429,7 @@ nf_nat_setup_info(struct nf_conn *ct,
srchash = hash_by_src(net,
&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
- lock = &nf_nat_locks[srchash % ARRAY_SIZE(nf_nat_locks)];
+ lock = &nf_nat_locks[srchash % CONNTRACK_LOCKS];
spin_lock_bh(lock);
hlist_add_head_rcu(&ct->nat_bysource,
&nf_nat_bysource[srchash]);
@@ -532,9 +532,9 @@ static void __nf_nat_cleanup_conntrack(struct nf_conn *ct)
unsigned int h;
h = hash_by_src(nf_ct_net(ct), &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
- spin_lock_bh(&nf_nat_locks[h % ARRAY_SIZE(nf_nat_locks)]);
+ spin_lock_bh(&nf_nat_locks[h % CONNTRACK_LOCKS]);
hlist_del_rcu(&ct->nat_bysource);
- spin_unlock_bh(&nf_nat_locks[h % ARRAY_SIZE(nf_nat_locks)]);
+ spin_unlock_bh(&nf_nat_locks[h % CONNTRACK_LOCKS]);
}
static int nf_nat_proto_clean(struct nf_conn *ct, void *data)
@@ -807,8 +807,8 @@ static int __init nf_nat_init(void)
/* Leave them the same for the moment. */
nf_nat_htable_size = nf_conntrack_htable_size;
- if (nf_nat_htable_size < ARRAY_SIZE(nf_nat_locks))
- nf_nat_htable_size = ARRAY_SIZE(nf_nat_locks);
+ if (nf_nat_htable_size < CONNTRACK_LOCKS)
+ nf_nat_htable_size = CONNTRACK_LOCKS;
nf_nat_bysource = nf_ct_alloc_hashtable(&nf_nat_htable_size, 0);
if (!nf_nat_bysource)
@@ -821,7 +821,7 @@ static int __init nf_nat_init(void)
return ret;
}
- for (i = 0; i < ARRAY_SIZE(nf_nat_locks); i++)
+ for (i = 0; i < CONNTRACK_LOCKS; i++)
spin_lock_init(&nf_nat_locks[i]);
nf_ct_helper_expectfn_register(&follow_master_nat);
--
2.1.4
^ permalink raw reply related
* [PATCH 0/2] Netfilter fixes for net
From: Pablo Neira Ayuso @ 2017-09-20 10:49 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
Hi David,
The following patchset contains two Netfilter fixes for your net tree,
they are:
1) Fix NAt compilation with UP, from Geert Uytterhoeven.
2) Fix incorrect number of entries when dumping a set, from
Vishwanath Pai.
You can pull these changes from:
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
Thanks!
----------------------------------------------------------------
The following changes since commit 2bd6bf03f4c1c59381d62c61d03f6cc3fe71f66e:
Linux 4.14-rc1 (2017-09-16 15:47:51 -0700)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git HEAD
for you to fetch changes up to 7f4f7dd4417d9efd038b14d39c70170db2e0baa0:
netfilter: ipset: ipset list may return wrong member count for set with timeout (2017-09-18 17:35:32 +0200)
----------------------------------------------------------------
Geert Uytterhoeven (1):
netfilter: nat: Do not use ARRAY_SIZE() on spinlocks to fix zero div
Vishwanath Pai (1):
netfilter: ipset: ipset list may return wrong member count for set with timeout
net/netfilter/ipset/ip_set_hash_gen.h | 14 +++++++++++++-
net/netfilter/nf_nat_core.c | 12 ++++++------
2 files changed, 19 insertions(+), 7 deletions(-)
^ 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