* [PATCH v2 net-next 0/2] tcp: simplify ack tx timestamps
From: Soheil Hassas Yeganeh @ 2016-04-28 3:38 UTC (permalink / raw)
To: davem, netdev
Cc: kafai, willemb, edumazet, ycheng, ncardwell,
Soheil Hassas Yeganeh
From: Soheil Hassas Yeganeh <soheil@google.com>
v2:
- Fully remove SKBTX_ACK_TSTAMP, as suggested by Willem de Bruijn.
This patch series aims at removing redundant checks and fields
for ack timestamps for TCP.
Soheil Hassas Yeganeh (2):
tcp: remove an unnecessary check in tcp_tx_timestamp
tcp: remove SKBTX_ACK_TSTAMP since it is redundant
include/linux/skbuff.h | 6 +-----
net/ipv4/tcp.c | 7 ++++---
net/ipv4/tcp_input.c | 3 +--
net/ipv4/tcp_output.c | 17 +++++++++++------
net/socket.c | 3 ---
5 files changed, 17 insertions(+), 19 deletions(-)
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply
* [PATCH v2 net-next 1/2] tcp: remove an unnecessary check in tcp_tx_timestamp
From: Soheil Hassas Yeganeh @ 2016-04-28 3:39 UTC (permalink / raw)
To: davem, netdev
Cc: kafai, willemb, edumazet, ycheng, ncardwell,
Soheil Hassas Yeganeh
In-Reply-To: <1461814741-848-1-git-send-email-soheil.kdev@gmail.com>
From: Soheil Hassas Yeganeh <soheil@google.com>
Remove the redundant check for sk->sk_tsflags in tcp_tx_timestamp.
tcp_tx_timestamp() receives the tsflags as a parameter. As a
result the "sk->sk_tsflags || tsflags" is redundant, since
tsflags already includes sk->sk_tsflags plus overrides from
control messages.
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
---
net/ipv4/tcp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 4d73858..3c542dc 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -430,7 +430,7 @@ EXPORT_SYMBOL(tcp_init_sock);
static void tcp_tx_timestamp(struct sock *sk, u16 tsflags, struct sk_buff *skb)
{
- if (sk->sk_tsflags || tsflags) {
+ if (tsflags) {
struct skb_shared_info *shinfo = skb_shinfo(skb);
struct tcp_skb_cb *tcb = TCP_SKB_CB(skb);
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH v2 net-next 2/2] tcp: remove SKBTX_ACK_TSTAMP since it is redundant
From: Soheil Hassas Yeganeh @ 2016-04-28 3:39 UTC (permalink / raw)
To: davem, netdev
Cc: kafai, willemb, edumazet, ycheng, ncardwell,
Soheil Hassas Yeganeh
In-Reply-To: <1461814741-848-1-git-send-email-soheil.kdev@gmail.com>
From: Soheil Hassas Yeganeh <soheil@google.com>
The SKBTX_ACK_TSTAMP flag is set in skb_shinfo->tx_flags when
the timestamp of the TCP acknowledgement should be reported on
error queue. Since accessing skb_shinfo is likely to incur a
cache-line miss at the time of receiving the ack, the
txstamp_ack bit was added in tcp_skb_cb, which is set iff
the SKBTX_ACK_TSTAMP flag is set for an skb. This makes
SKBTX_ACK_TSTAMP flag redundant.
Remove the SKBTX_ACK_TSTAMP and instead use the txstamp_ack bit
everywhere.
Note that this frees one bit in shinfo->tx_flags.
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Suggested-by: Willem de Bruijn <willemb@google.com>
---
include/linux/skbuff.h | 6 +-----
net/ipv4/tcp.c | 5 +++--
net/ipv4/tcp_input.c | 3 +--
net/ipv4/tcp_output.c | 17 +++++++++++------
net/socket.c | 3 ---
5 files changed, 16 insertions(+), 18 deletions(-)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index da0ace3..ae30555 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -382,14 +382,10 @@ enum {
/* generate software time stamp when entering packet scheduling */
SKBTX_SCHED_TSTAMP = 1 << 6,
-
- /* generate software timestamp on peer data acknowledgment */
- SKBTX_ACK_TSTAMP = 1 << 7,
};
#define SKBTX_ANY_SW_TSTAMP (SKBTX_SW_TSTAMP | \
- SKBTX_SCHED_TSTAMP | \
- SKBTX_ACK_TSTAMP)
+ SKBTX_SCHED_TSTAMP)
#define SKBTX_ANY_TSTAMP (SKBTX_HW_TSTAMP | SKBTX_ANY_SW_TSTAMP)
/*
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 3c542dc..8e05eb6 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -435,9 +435,10 @@ static void tcp_tx_timestamp(struct sock *sk, u16 tsflags, struct sk_buff *skb)
struct tcp_skb_cb *tcb = TCP_SKB_CB(skb);
sock_tx_timestamp(sk, tsflags, &shinfo->tx_flags);
- if (shinfo->tx_flags & SKBTX_ANY_TSTAMP)
+ if (tsflags & SOF_TIMESTAMPING_TX_ACK)
+ tcb->txstamp_ack = 1;
+ if (tsflags & SOF_TIMESTAMPING_TX_RECORD_MASK)
shinfo->tskey = TCP_SKB_CB(skb)->seq + skb->len - 1;
- tcb->txstamp_ack = !!(shinfo->tx_flags & SKBTX_ACK_TSTAMP);
}
}
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 967520d..2f3fd92 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3087,8 +3087,7 @@ static void tcp_ack_tstamp(struct sock *sk, struct sk_buff *skb,
return;
shinfo = skb_shinfo(skb);
- if ((shinfo->tx_flags & SKBTX_ACK_TSTAMP) &&
- !before(shinfo->tskey, prior_snd_una) &&
+ if (!before(shinfo->tskey, prior_snd_una) &&
before(shinfo->tskey, tcp_sk(sk)->snd_una))
__skb_tstamp_tx(skb, NULL, sk, SCM_TSTAMP_ACK);
}
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 9d3b4b3..ace183c 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1111,11 +1111,17 @@ static void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int de
tcp_verify_left_out(tp);
}
+static bool tcp_has_tx_tstamp(const struct sk_buff *skb)
+{
+ return TCP_SKB_CB(skb)->txstamp_ack ||
+ (skb_shinfo(skb)->tx_flags & SKBTX_ANY_TSTAMP);
+}
+
static void tcp_fragment_tstamp(struct sk_buff *skb, struct sk_buff *skb2)
{
struct skb_shared_info *shinfo = skb_shinfo(skb);
- if (unlikely(shinfo->tx_flags & SKBTX_ANY_TSTAMP) &&
+ if (unlikely(tcp_has_tx_tstamp(skb)) &&
!before(shinfo->tskey, TCP_SKB_CB(skb2)->seq)) {
struct skb_shared_info *shinfo2 = skb_shinfo(skb2);
u8 tsflags = shinfo->tx_flags & SKBTX_ANY_TSTAMP;
@@ -2446,13 +2452,12 @@ u32 __tcp_select_window(struct sock *sk)
void tcp_skb_collapse_tstamp(struct sk_buff *skb,
const struct sk_buff *next_skb)
{
- const struct skb_shared_info *next_shinfo = skb_shinfo(next_skb);
- u8 tsflags = next_shinfo->tx_flags & SKBTX_ANY_TSTAMP;
-
- if (unlikely(tsflags)) {
+ if (unlikely(tcp_has_tx_tstamp(next_skb))) {
+ const struct skb_shared_info *next_shinfo =
+ skb_shinfo(next_skb);
struct skb_shared_info *shinfo = skb_shinfo(skb);
- shinfo->tx_flags |= tsflags;
+ shinfo->tx_flags |= next_shinfo->tx_flags & SKBTX_ANY_TSTAMP;
shinfo->tskey = next_shinfo->tskey;
TCP_SKB_CB(skb)->txstamp_ack |=
TCP_SKB_CB(next_skb)->txstamp_ack;
diff --git a/net/socket.c b/net/socket.c
index 5dbb0bb..7789d79 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -600,9 +600,6 @@ void __sock_tx_timestamp(__u16 tsflags, __u8 *tx_flags)
if (tsflags & SOF_TIMESTAMPING_TX_SCHED)
flags |= SKBTX_SCHED_TSTAMP;
- if (tsflags & SOF_TIMESTAMPING_TX_ACK)
- flags |= SKBTX_ACK_TSTAMP;
-
*tx_flags = flags;
}
EXPORT_SYMBOL(__sock_tx_timestamp);
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* Re: [RFC PATCH 4/5] bnxt: Add support for segmentation of tunnels with outer checksums
From: Michael Chan @ 2016-04-28 4:32 UTC (permalink / raw)
To: Alexander Duyck
Cc: Alexander Duyck, eugenia, Bruce W Allan, Saeed Mahameed, Netdev,
intel-wired-lan, Ariel Elior, Michael Chan
In-Reply-To: <CAKgT0UffVkQWWqgwYZy-AXEq=kofKiRx2D9pnu5nBQD+uSLS+g@mail.gmail.com>
On Wed, Apr 27, 2016 at 8:21 AM, Alexander Duyck
<alexander.duyck@gmail.com> wrote:
> On Tue, Apr 26, 2016 at 10:55 PM, Michael Chan
> <michael.chan@broadcom.com> wrote:
>> On Tue, Apr 19, 2016 at 12:06 PM, Alexander Duyck <aduyck@mirantis.com> wrote:
>>> This patch assumes that the bnxt hardware will ignore existing IPv4/v6
>>> header fields for length and checksum as well as the length and checksum
>>> fields for outer UDP and GRE headers.
>>>
>>> I have no means of testing this as I do not have any bnx2x hardware but
>>> thought I would submit it as an RFC to see if anyone out there wants to
>>> test this and see if this does in fact enable this functionality allowing
>>> us to to segment tunneled frames that have an outer checksum.
>>>
>>> Signed-off-by: Alexander Duyck <aduyck@mirantis.com>
>>
>> Hi Alex, I just did a very quick test of this patch on our bnxt
>> hardware and it seemed to work.
>>
>> I created a vxlan endpoint with udpcsum enabled and I saw TSO packets
>> getting through. I've verified that our hardware can be programmed to
>> either ignore outer UDP checksum or to calculate it. Current default
>> is to ignore ipv4 UDP checksum and calculate ipv6 UDP checksum.
>> Thanks.
>
> Are you saying you can natively support UDP tunnel with outer checksum
> offload then?
Yes. Calculate or ignore the outer UDP checksum.
>
> I'm just trying to sort out if you actually need to have the partial
> segmentation offload support or if we can handle it in hardware. Also
> is there any documentation you could point me to that might help to
> clarify what the hardware does/doesn't support so that I could improve
> upon this patch in order to make sure we are getting the most bang for
> the buck in terms of the features that can be offloaded by hardware?
No public documentation yet. I think the plan is to publish the
programmer's reference on our website at some point in the future.
^ permalink raw reply
* Re: [PATCH v2 net-next 1/2] tcp: remove an unnecessary check in tcp_tx_timestamp
From: Eric Dumazet @ 2016-04-28 4:35 UTC (permalink / raw)
To: Soheil Hassas Yeganeh
Cc: davem, netdev, kafai, willemb, edumazet, ycheng, ncardwell,
Soheil Hassas Yeganeh
In-Reply-To: <1461814741-848-2-git-send-email-soheil.kdev@gmail.com>
On Wed, 2016-04-27 at 23:39 -0400, Soheil Hassas Yeganeh wrote:
> From: Soheil Hassas Yeganeh <soheil@google.com>
>
> Remove the redundant check for sk->sk_tsflags in tcp_tx_timestamp.
>
> tcp_tx_timestamp() receives the tsflags as a parameter. As a
> result the "sk->sk_tsflags || tsflags" is redundant, since
> tsflags already includes sk->sk_tsflags plus overrides from
> control messages.
>
> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
> ---
Acked-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: [PATCH v2 net-next 2/2] tcp: remove SKBTX_ACK_TSTAMP since it is redundant
From: Eric Dumazet @ 2016-04-28 4:37 UTC (permalink / raw)
To: Soheil Hassas Yeganeh
Cc: davem, netdev, kafai, willemb, edumazet, ycheng, ncardwell,
Soheil Hassas Yeganeh
In-Reply-To: <1461814741-848-3-git-send-email-soheil.kdev@gmail.com>
On Wed, 2016-04-27 at 23:39 -0400, Soheil Hassas Yeganeh wrote:
> From: Soheil Hassas Yeganeh <soheil@google.com>
>
> The SKBTX_ACK_TSTAMP flag is set in skb_shinfo->tx_flags when
> the timestamp of the TCP acknowledgement should be reported on
> error queue. Since accessing skb_shinfo is likely to incur a
> cache-line miss at the time of receiving the ack, the
> txstamp_ack bit was added in tcp_skb_cb, which is set iff
> the SKBTX_ACK_TSTAMP flag is set for an skb. This makes
> SKBTX_ACK_TSTAMP flag redundant.
>
> Remove the SKBTX_ACK_TSTAMP and instead use the txstamp_ack bit
> everywhere.
>
> Note that this frees one bit in shinfo->tx_flags.
>
> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
> Acked-by: Martin KaFai Lau <kafai@fb.com>
> Suggested-by: Willem de Bruijn <willemb@google.com>
> ---
Acked-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: iproute2: bash completion function for tc
From: Stephen Hemminger @ 2016-04-28 5:13 UTC (permalink / raw)
To: Alexei Starovoitov
Cc: Quentin Monnet, Jamal Hadi Salim, Vincent Jardin, netdev
In-Reply-To: <20160428031923.GA76700@ast-mbp.thefacebook.com>
On Wed, 27 Apr 2016 20:19:26 -0700
Alexei Starovoitov <alexei.starovoitov@gmail.com> wrote:
> On Tue, Apr 26, 2016 at 09:28:17AM +0200, Quentin Monnet wrote:
> > Hi Jamal, Stephen,
> >
> > I searched for a function providing auto-completion for `tc` utility in
> > bash, but I found none. So I have created one, and I would like share it
> > with the community. It is available here:
> > https://github.com/6WIND/tc_bash-completion/blob/master/tc
> > I would like to make it easily available to tc users, so here is a
> > twofold request:
> >
> > * I do not know where to submit the code. Should I submit here on netdev
> > for inclusion in iproute2 package, or rather to the bash-completion
> > repository on GitHub? I feel like it would receive better feedback and
> > updates if pushed to iproute2. Could you please provide some advice here?
> > * The completion for `tc` seems to work well; I have tested it with many
> > commands, but I am no tc expert, and there are probably some cases where
> > the completion fails to propose the correct choices. I would be really
> > interested in any feedback/bug reports that you, or anyone on this list
> > who uses tc, could provide.
>
> that looks very interesting.
> I think making it a part of iproute2 is a good thing.
> How about installing it into /etc/iproute2/ ?
> Stephen, any comments?
>
I am ok with keeping it in the repository.
But it would need to be installed in the standard bash directory,
is that distro dependent?
^ permalink raw reply
* Re: [PATCH net-next 2/7] net: rtnetlink: allow only one idx saving stats attribute
From: Roopa Prabhu @ 2016-04-28 5:18 UTC (permalink / raw)
To: Nikolay Aleksandrov; +Cc: netdev, davem, stephen, jhs
In-Reply-To: <1461773902-13528-3-git-send-email-nikolay@cumulusnetworks.com>
On 4/27/16, 9:18 AM, Nikolay Aleksandrov wrote:
> We can't allow more than one stats attribute which uses the local idx
> since the result will be a mess. This is a simple check to make sure
> only one is being used at a time. Later when the filter_mask's 32 bits
> are over we can switch to a bitmap.
>
> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
> ---
> include/net/rtnetlink.h | 6 ++++++
> net/core/rtnetlink.c | 17 +++++++++++++++--
> 2 files changed, 21 insertions(+), 2 deletions(-)
>
> diff --git a/include/net/rtnetlink.h b/include/net/rtnetlink.h
> index 2f87c1ba13de..3f3b0b1b8722 100644
> --- a/include/net/rtnetlink.h
> +++ b/include/net/rtnetlink.h
> @@ -150,4 +150,10 @@ int rtnl_nla_parse_ifla(struct nlattr **tb, const struct nlattr *head, int len);
>
> #define MODULE_ALIAS_RTNL_LINK(kind) MODULE_ALIAS("rtnl-link-" kind)
>
> +/* at most one attribute which can save a local idx is allowed to be set
> + * IFLA_STATS_IDX_ATTR_MASK has all the idx saving attributes set and is
> + * used to check if more than one is being requested
> + */
> +#define IFLA_STATS_IDX_ATTR_MASK 0
> +
> #endif
> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index aeb2fa9b1cda..ea03b6cd3d3c 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -3512,7 +3512,7 @@ static int rtnl_stats_get(struct sk_buff *skb, struct nlmsghdr *nlh)
> struct if_stats_msg *ifsm;
> struct net_device *dev = NULL;
> struct sk_buff *nskb;
> - u32 filter_mask;
> + u32 filter_mask, lidx_filter;
> int lidx = 0;
> int err;
>
> @@ -3529,6 +3529,14 @@ static int rtnl_stats_get(struct sk_buff *skb, struct nlmsghdr *nlh)
> if (!filter_mask)
> return -EINVAL;
>
> + /* only one attribute which can save a local idx is allowed at a time
> + * even though rtnl_stats_get doesn't save the lidx, we need to be
> + * consistent with the dump side and error out
> + */
> + lidx_filter = filter_mask & IFLA_STATS_IDX_ATTR_MASK;
> + if (lidx_filter && !is_power_of_2(lidx_filter))
> + return -EINVAL;
> +
> nskb = nlmsg_new(if_nlmsg_stats_size(dev, filter_mask), GFP_KERNEL);
> if (!nskb)
> return -ENOBUFS;
> @@ -3556,7 +3564,7 @@ static int rtnl_stats_dump(struct sk_buff *skb, struct netlink_callback *cb)
> struct net_device *dev;
> struct hlist_head *head;
> unsigned int flags = NLM_F_MULTI;
> - u32 filter_mask = 0;
> + u32 filter_mask = 0, lidx_filter;
> int err;
>
> s_h = cb->args[0];
> @@ -3570,6 +3578,11 @@ static int rtnl_stats_dump(struct sk_buff *skb, struct netlink_callback *cb)
> if (!filter_mask)
> return -EINVAL;
>
> + /* only one attribute which can save a local idx is allowed at a time */
> + lidx_filter = filter_mask & IFLA_STATS_IDX_ATTR_MASK;
> + if (lidx_filter && !is_power_of_2(lidx_filter))
> + return -EINVAL;
> +
>
instead of introducing the restriction at this level, is it possible to use two args for this
like below and avoid the restriction ?
cb->args[2] = current filter being processed
cb->args[3] = private filter idx (your lidx)
^ permalink raw reply
* [PATCH net-next 0/6] net: make TCP preemptible
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
Most of TCP stack assumed it was running from BH handler.
This is great for most things, as TCP behavior is very sensitive
to scheduling artifacts.
However, the prequeue and backlog processing are problematic,
as they need to be flushed with BH being blocked.
To cope with modern needs, TCP sockets have big sk_rcvbuf values,
in the order of 16 MB.
This means that backlog can hold thousands of packets, and things
like TCP coalescing or collapsing on this amount of packets can
lead to insane latency spikes, since BH are blocked for too long.
It is time to make UDP/TCP stacks preemptible.
Note that fast path still runs from BH handler.
Eric Dumazet (6):
tcp: do not assume TCP code is non preemptible
tcp: do not block bh during prequeue processing
dccp: do not assume DCCP code is non preemptible
udp: prepare for non BH masking at backlog processing
sctp: prepare for socket backlog behavior change
net: do not block BH while processing socket backlog
net/core/sock.c | 22 +++------
net/dccp/input.c | 2 +-
net/dccp/ipv4.c | 4 +-
net/dccp/ipv6.c | 4 +-
net/dccp/options.c | 2 +-
net/ipv4/tcp.c | 6 +--
net/ipv4/tcp_cdg.c | 20 ++++----
net/ipv4/tcp_cubic.c | 20 ++++----
net/ipv4/tcp_fastopen.c | 12 ++---
net/ipv4/tcp_input.c | 126 +++++++++++++++++++----------------------------
net/ipv4/tcp_ipv4.c | 14 ++++--
net/ipv4/tcp_minisocks.c | 2 +-
net/ipv4/tcp_output.c | 7 ++-
net/ipv4/tcp_recovery.c | 4 +-
net/ipv4/tcp_timer.c | 10 ++--
net/ipv4/udp.c | 4 +-
net/ipv6/tcp_ipv6.c | 12 ++---
net/ipv6/udp.c | 4 +-
net/sctp/inqueue.c | 2 +
19 files changed, 124 insertions(+), 153 deletions(-)
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply
* [PATCH net-next 2/6] tcp: do not block bh during prequeue processing
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
AFAIK, nothing in current TCP stack absolutely wants BH
being disabled once socket is owned by a thread running in
process context.
As mentioned in my prior patch ("tcp: give prequeue mode some care"),
processing a batch of packets might take time, better not block BH
at all.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/ipv4/tcp.c | 4 ----
net/ipv4/tcp_input.c | 30 ++----------------------------
2 files changed, 2 insertions(+), 32 deletions(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 7f51389814e6..f8856b76f941 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1445,12 +1445,8 @@ static void tcp_prequeue_process(struct sock *sk)
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPPREQUEUED);
- /* RX process wants to run with disabled BHs, though it is not
- * necessary */
- local_bh_disable();
while ((skb = __skb_dequeue(&tp->ucopy.prequeue)) != NULL)
sk_backlog_rcv(sk, skb);
- local_bh_enable();
/* Clear memory counter. */
tp->ucopy.memory = 0;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 0eb31df8edfa..44e0f9f15f32 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4608,14 +4608,12 @@ static void tcp_data_queue(struct sock *sk, struct sk_buff *skb)
__set_current_state(TASK_RUNNING);
- local_bh_enable();
if (!skb_copy_datagram_msg(skb, 0, tp->ucopy.msg, chunk)) {
tp->ucopy.len -= chunk;
tp->copied_seq += chunk;
eaten = (chunk == skb->len);
tcp_rcv_space_adjust(sk);
}
- local_bh_disable();
}
if (eaten <= 0) {
@@ -5131,7 +5129,6 @@ static int tcp_copy_to_iovec(struct sock *sk, struct sk_buff *skb, int hlen)
int chunk = skb->len - hlen;
int err;
- local_bh_enable();
if (skb_csum_unnecessary(skb))
err = skb_copy_datagram_msg(skb, hlen, tp->ucopy.msg, chunk);
else
@@ -5143,32 +5140,9 @@ static int tcp_copy_to_iovec(struct sock *sk, struct sk_buff *skb, int hlen)
tcp_rcv_space_adjust(sk);
}
- local_bh_disable();
return err;
}
-static __sum16 __tcp_checksum_complete_user(struct sock *sk,
- struct sk_buff *skb)
-{
- __sum16 result;
-
- if (sock_owned_by_user(sk)) {
- local_bh_enable();
- result = __tcp_checksum_complete(skb);
- local_bh_disable();
- } else {
- result = __tcp_checksum_complete(skb);
- }
- return result;
-}
-
-static inline bool tcp_checksum_complete_user(struct sock *sk,
- struct sk_buff *skb)
-{
- return !skb_csum_unnecessary(skb) &&
- __tcp_checksum_complete_user(sk, skb);
-}
-
/* Does PAWS and seqno based validation of an incoming segment, flags will
* play significant role here.
*/
@@ -5383,7 +5357,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
}
}
if (!eaten) {
- if (tcp_checksum_complete_user(sk, skb))
+ if (tcp_checksum_complete(skb))
goto csum_error;
if ((int)skb->truesize > sk->sk_forward_alloc)
@@ -5427,7 +5401,7 @@ no_ack:
}
slow_path:
- if (len < (th->doff << 2) || tcp_checksum_complete_user(sk, skb))
+ if (len < (th->doff << 2) || tcp_checksum_complete(skb))
goto csum_error;
if (!th->ack && !th->rst && !th->syn)
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH net-next 3/6] dccp: do not assume DCCP code is non preemptible
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
DCCP uses the generic backlog code, and this will soon
be changed to not disable BH when protocol is called back.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/dccp/input.c | 2 +-
net/dccp/ipv4.c | 4 ++--
net/dccp/ipv6.c | 4 ++--
net/dccp/options.c | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/net/dccp/input.c b/net/dccp/input.c
index 2437ecc13b82..ba347184bda9 100644
--- a/net/dccp/input.c
+++ b/net/dccp/input.c
@@ -359,7 +359,7 @@ send_sync:
goto discard;
}
- __DCCP_INC_STATS(DCCP_MIB_INERRS);
+ DCCP_INC_STATS(DCCP_MIB_INERRS);
discard:
__kfree_skb(skb);
return 0;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index a8164272e0f4..5c7e413a3ae4 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -533,8 +533,8 @@ static void dccp_v4_ctl_send_reset(const struct sock *sk, struct sk_buff *rxskb)
bh_unlock_sock(ctl_sk);
if (net_xmit_eval(err) == 0) {
- __DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
- __DCCP_INC_STATS(DCCP_MIB_OUTRSTS);
+ DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
+ DCCP_INC_STATS(DCCP_MIB_OUTRSTS);
}
out:
dst_release(dst);
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 0f4eb4ea57a5..d176f4e66369 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -277,8 +277,8 @@ static void dccp_v6_ctl_send_reset(const struct sock *sk, struct sk_buff *rxskb)
if (!IS_ERR(dst)) {
skb_dst_set(skb, dst);
ip6_xmit(ctl_sk, skb, &fl6, NULL, 0);
- __DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
- __DCCP_INC_STATS(DCCP_MIB_OUTRSTS);
+ DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
+ DCCP_INC_STATS(DCCP_MIB_OUTRSTS);
return;
}
diff --git a/net/dccp/options.c b/net/dccp/options.c
index b82b7ee9a1d2..74d29c56c367 100644
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -253,7 +253,7 @@ out_nonsensical_length:
return 0;
out_invalid_option:
- __DCCP_INC_STATS(DCCP_MIB_INVALIDOPT);
+ DCCP_INC_STATS(DCCP_MIB_INVALIDOPT);
rc = DCCP_RESET_CODE_OPTION_ERROR;
out_featneg_failed:
DCCP_WARN("DCCP(%p): Option %d (len=%d) error=%u\n", sk, opt, len, rc);
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH net-next 1/6] tcp: do not assume TCP code is non preemptible
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
We want to to make TCP stack preemptible, as draining prequeue
and backlog queues can take lot of time.
Many SNMP updates were assuming that BH (and preemption) was disabled.
Need to convert some __NET_INC_STATS() calls to NET_INC_STATS()
and some __TCP_INC_STATS() to TCP_INC_STATS()
Before using this_cpu_ptr(net->ipv4.tcp_sk) in tcp_v4_send_reset()
and tcp_v4_send_ack(), we add an explicit preempt disabled section.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/ipv4/tcp.c | 2 +-
net/ipv4/tcp_cdg.c | 20 +++++-----
net/ipv4/tcp_cubic.c | 20 +++++-----
net/ipv4/tcp_fastopen.c | 12 +++---
net/ipv4/tcp_input.c | 96 ++++++++++++++++++++++++------------------------
net/ipv4/tcp_ipv4.c | 14 ++++---
net/ipv4/tcp_minisocks.c | 2 +-
net/ipv4/tcp_output.c | 7 ++--
net/ipv4/tcp_recovery.c | 4 +-
net/ipv4/tcp_timer.c | 10 +++--
net/ipv6/tcp_ipv6.c | 12 +++---
11 files changed, 102 insertions(+), 97 deletions(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 040f35e7efe0..7f51389814e6 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -3091,7 +3091,7 @@ void tcp_done(struct sock *sk)
struct request_sock *req = tcp_sk(sk)->fastopen_rsk;
if (sk->sk_state == TCP_SYN_SENT || sk->sk_state == TCP_SYN_RECV)
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
tcp_set_state(sk, TCP_CLOSE);
tcp_clear_xmit_timers(sk);
diff --git a/net/ipv4/tcp_cdg.c b/net/ipv4/tcp_cdg.c
index 3c00208c37f4..4e3007845888 100644
--- a/net/ipv4/tcp_cdg.c
+++ b/net/ipv4/tcp_cdg.c
@@ -155,11 +155,11 @@ static void tcp_cdg_hystart_update(struct sock *sk)
ca->last_ack = now_us;
if (after(now_us, ca->round_start + base_owd)) {
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTTRAINDETECT);
- __NET_ADD_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTTRAINCWND,
- tp->snd_cwnd);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTTRAINDETECT);
+ NET_ADD_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTTRAINCWND,
+ pp>>sn__cwdd);
tp->snd_ssthresh = tp->snd_cwnd;
return;
}
@@ -174,11 +174,11 @@ static void tcp_cdg_hystart_update(struct sock *sk)
125U);
if (ca->rtt.min > thresh) {
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTDELAYDETECT);
- __NET_ADD_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTDELAYCWND,
- tp->snd_cwnd);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTDELAYDETECT);
+ NET_ADD_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTDELAYCWND,
+ tp->snd_cwnd);
tp->snd_ssthresh = tp->snd_cwnd;
}
}
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index 59155af9de5d..0ce946e395e1 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -402,11 +402,11 @@ static void hystart_update(struct sock *sk, u32 delay)
ca->last_ack = now;
if ((s32)(now - ca->round_start) > ca->delay_min >> 4) {
ca->found |= HYSTART_ACK_TRAIN;
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTTRAINDETECT);
- __NET_ADD_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTTRAINCWND,
- tp->snd_cwnd);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTTRAINDETECT);
+ NET_ADD_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTTRAINCWND,
+ tp->snd_cwnd);
tp->snd_ssthresh = tp->snd_cwnd;
}
}
@@ -423,11 +423,11 @@ static void hystart_update(struct sock *sk, u32 delay)
if (ca->curr_rtt > ca->delay_min +
HYSTART_DELAY_THRESH(ca->delay_min >> 3)) {
ca->found |= HYSTART_DELAY;
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTDELAYDETECT);
- __NET_ADD_STATS(sock_net(sk),
- LINUX_MIB_TCPHYSTARTDELAYCWND,
- tp->snd_cwnd);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTDELAYDETECT);
+ NET_ADD_STATS(sock_net(sk),
+ LINUX_MIB_TCPHYSTARTDELAYCWND,
+ tp->snd_cwnd);
tp->snd_ssthresh = tp->snd_cwnd;
}
}
diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
index a1498d507e42..54d9f9b0120f 100644
--- a/net/ipv4/tcp_fastopen.c
+++ b/net/ipv4/tcp_fastopen.c
@@ -255,9 +255,9 @@ static bool tcp_fastopen_queue_check(struct sock *sk)
spin_lock(&fastopenq->lock);
req1 = fastopenq->rskq_rst_head;
if (!req1 || time_after(req1->rsk_timer.expires, jiffies)) {
- spin_unlock(&fastopenq->lock);
__NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPFASTOPENLISTENOVERFLOW);
+ spin_unlock(&fastopenq->lock);
return false;
}
fastopenq->rskq_rst_head = req1->dl_next;
@@ -282,7 +282,7 @@ struct sock *tcp_try_fastopen(struct sock *sk, struct sk_buff *skb,
struct sock *child;
if (foc->len == 0) /* Client requests a cookie */
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENCOOKIEREQD);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENCOOKIEREQD);
if (!((sysctl_tcp_fastopen & TFO_SERVER_ENABLE) &&
(syn_data || foc->len >= 0) &&
@@ -311,13 +311,13 @@ fastopen:
child = tcp_fastopen_create_child(sk, skb, dst, req);
if (child) {
foc->len = -1;
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPFASTOPENPASSIVE);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPFASTOPENPASSIVE);
return child;
}
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENPASSIVEFAIL);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENPASSIVEFAIL);
} else if (foc->len > 0) /* Client presents an invalid cookie */
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENPASSIVEFAIL);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFASTOPENPASSIVEFAIL);
valid_foc.exp = foc->exp;
*foc = valid_foc;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 0d5239c283cb..0eb31df8edfa 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -869,7 +869,7 @@ static void tcp_update_reordering(struct sock *sk, const int metric,
else
mib_idx = LINUX_MIB_TCPSACKREORDER;
- __NET_INC_STATS(sock_net(sk), mib_idx);
+ NET_INC_STATS(sock_net(sk), mib_idx);
#if FASTRETRANS_DEBUG > 1
pr_debug("Disorder%d %d %u f%u s%u rr%d\n",
tp->rx_opt.sack_ok, inet_csk(sk)->icsk_ca_state,
@@ -1062,7 +1062,7 @@ static bool tcp_check_dsack(struct sock *sk, const struct sk_buff *ack_skb,
if (before(start_seq_0, TCP_SKB_CB(ack_skb)->ack_seq)) {
dup_sack = true;
tcp_dsack_seen(tp);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPDSACKRECV);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPDSACKRECV);
} else if (num_sacks > 1) {
u32 end_seq_1 = get_unaligned_be32(&sp[1].end_seq);
u32 start_seq_1 = get_unaligned_be32(&sp[1].start_seq);
@@ -1071,7 +1071,7 @@ static bool tcp_check_dsack(struct sock *sk, const struct sk_buff *ack_skb,
!before(start_seq_0, start_seq_1)) {
dup_sack = true;
tcp_dsack_seen(tp);
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPDSACKOFORECV);
}
}
@@ -1289,7 +1289,7 @@ static bool tcp_shifted_skb(struct sock *sk, struct sk_buff *skb,
if (skb->len > 0) {
BUG_ON(!tcp_skb_pcount(skb));
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKSHIFTED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKSHIFTED);
return false;
}
@@ -1313,7 +1313,7 @@ static bool tcp_shifted_skb(struct sock *sk, struct sk_buff *skb,
tcp_unlink_write_queue(skb, sk);
sk_wmem_free_skb(sk, skb);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKMERGED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKMERGED);
return true;
}
@@ -1469,7 +1469,7 @@ noop:
return skb;
fallback:
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKSHIFTFALLBACK);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_SACKSHIFTFALLBACK);
return NULL;
}
@@ -1657,7 +1657,7 @@ tcp_sacktag_write_queue(struct sock *sk, const struct sk_buff *ack_skb,
mib_idx = LINUX_MIB_TCPSACKDISCARD;
}
- __NET_INC_STATS(sock_net(sk), mib_idx);
+ NET_INC_STATS(sock_net(sk), mib_idx);
if (i == 0)
first_sack_index = -1;
continue;
@@ -1909,7 +1909,7 @@ void tcp_enter_loss(struct sock *sk)
skb = tcp_write_queue_head(sk);
is_reneg = skb && (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED);
if (is_reneg) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSACKRENEGING);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSACKRENEGING);
tp->sacked_out = 0;
tp->fackets_out = 0;
}
@@ -2395,7 +2395,7 @@ static bool tcp_try_undo_recovery(struct sock *sk)
else
mib_idx = LINUX_MIB_TCPFULLUNDO;
- __NET_INC_STATS(sock_net(sk), mib_idx);
+ NET_INC_STATS(sock_net(sk), mib_idx);
}
if (tp->snd_una == tp->high_seq && tcp_is_reno(tp)) {
/* Hold old state until something *above* high_seq
@@ -2417,7 +2417,7 @@ static bool tcp_try_undo_dsack(struct sock *sk)
if (tp->undo_marker && !tp->undo_retrans) {
DBGUNDO(sk, "D-SACK");
tcp_undo_cwnd_reduction(sk, false);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPDSACKUNDO);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPDSACKUNDO);
return true;
}
return false;
@@ -2432,9 +2432,9 @@ static bool tcp_try_undo_loss(struct sock *sk, bool frto_undo)
tcp_undo_cwnd_reduction(sk, true);
DBGUNDO(sk, "partial loss");
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSUNDO);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSUNDO);
if (frto_undo)
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPSPURIOUSRTOS);
inet_csk(sk)->icsk_retransmits = 0;
if (frto_undo || tcp_is_sack(tp))
@@ -2559,7 +2559,7 @@ static void tcp_mtup_probe_failed(struct sock *sk)
icsk->icsk_mtup.search_high = icsk->icsk_mtup.probe_size - 1;
icsk->icsk_mtup.probe_size = 0;
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMTUPFAIL);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMTUPFAIL);
}
static void tcp_mtup_probe_success(struct sock *sk)
@@ -2579,7 +2579,7 @@ static void tcp_mtup_probe_success(struct sock *sk)
icsk->icsk_mtup.search_low = icsk->icsk_mtup.probe_size;
icsk->icsk_mtup.probe_size = 0;
tcp_sync_mss(sk, icsk->icsk_pmtu_cookie);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMTUPSUCCESS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMTUPSUCCESS);
}
/* Do a simple retransmit without using the backoff mechanisms in
@@ -2643,7 +2643,7 @@ static void tcp_enter_recovery(struct sock *sk, bool ece_ack)
else
mib_idx = LINUX_MIB_TCPSACKRECOVERY;
- __NET_INC_STATS(sock_net(sk), mib_idx);
+ NET_INC_STATS(sock_net(sk), mib_idx);
tp->prior_ssthresh = 0;
tcp_init_undo(tp);
@@ -2736,7 +2736,7 @@ static bool tcp_try_undo_partial(struct sock *sk, const int acked)
DBGUNDO(sk, "partial recovery");
tcp_undo_cwnd_reduction(sk, true);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPPARTIALUNDO);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPPARTIALUNDO);
tcp_try_keep_open(sk);
return true;
}
@@ -3431,7 +3431,7 @@ bool tcp_oow_rate_limited(struct net *net, const struct sk_buff *skb,
s32 elapsed = (s32)(tcp_time_stamp - *last_oow_ack_time);
if (0 <= elapsed && elapsed < sysctl_tcp_invalid_ratelimit) {
- __NET_INC_STATS(net, mib_idx);
+ NET_INC_STATS(net, mib_idx);
return true; /* rate-limited: don't send yet! */
}
}
@@ -3464,7 +3464,7 @@ static void tcp_send_challenge_ack(struct sock *sk, const struct sk_buff *skb)
challenge_count = 0;
}
if (++challenge_count <= sysctl_tcp_challenge_ack_limit) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPCHALLENGEACK);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPCHALLENGEACK);
tcp_send_ack(sk);
}
}
@@ -3513,7 +3513,7 @@ static void tcp_process_tlp_ack(struct sock *sk, u32 ack, int flag)
tcp_set_ca_state(sk, TCP_CA_CWR);
tcp_end_cwnd_reduction(sk);
tcp_try_keep_open(sk);
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPLOSSPROBERECOVERY);
} else if (!(flag & (FLAG_SND_UNA_ADVANCED |
FLAG_NOT_DUP | FLAG_DATA_SACKED))) {
@@ -3618,14 +3618,14 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
tcp_in_ack_event(sk, CA_ACK_WIN_UPDATE);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPHPACKS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPHPACKS);
} else {
u32 ack_ev_flags = CA_ACK_SLOWPATH;
if (ack_seq != TCP_SKB_CB(skb)->end_seq)
flag |= FLAG_DATA;
else
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPPUREACKS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPPUREACKS);
flag |= tcp_ack_update_window(sk, skb, ack, ack_seq);
@@ -4128,7 +4128,7 @@ static void tcp_dsack_set(struct sock *sk, u32 seq, u32 end_seq)
else
mib_idx = LINUX_MIB_TCPDSACKOFOSENT;
- __NET_INC_STATS(sock_net(sk), mib_idx);
+ NET_INC_STATS(sock_net(sk), mib_idx);
tp->rx_opt.dsack = 1;
tp->duplicate_sack[0].start_seq = seq;
@@ -4152,7 +4152,7 @@ static void tcp_send_dupack(struct sock *sk, const struct sk_buff *skb)
if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq &&
before(TCP_SKB_CB(skb)->seq, tp->rcv_nxt)) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKLOST);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKLOST);
tcp_enter_quickack_mode(sk);
if (tcp_is_sack(tp) && sysctl_tcp_dsack) {
@@ -4302,7 +4302,7 @@ static bool tcp_try_coalesce(struct sock *sk,
atomic_add(delta, &sk->sk_rmem_alloc);
sk_mem_charge(sk, delta);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPRCVCOALESCE);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPRCVCOALESCE);
TCP_SKB_CB(to)->end_seq = TCP_SKB_CB(from)->end_seq;
TCP_SKB_CB(to)->ack_seq = TCP_SKB_CB(from)->ack_seq;
TCP_SKB_CB(to)->tcp_flags |= TCP_SKB_CB(from)->tcp_flags;
@@ -4390,7 +4390,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
tcp_ecn_check_ce(tp, skb);
if (unlikely(tcp_try_rmem_schedule(sk, skb, skb->truesize))) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFODROP);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFODROP);
tcp_drop(sk, skb);
return;
}
@@ -4399,7 +4399,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
tp->pred_flags = 0;
inet_csk_schedule_ack(sk);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOQUEUE);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOQUEUE);
SOCK_DEBUG(sk, "out of order segment: rcv_next %X seq %X - %X\n",
tp->rcv_nxt, TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq);
@@ -4454,7 +4454,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
if (skb1 && before(seq, TCP_SKB_CB(skb1)->end_seq)) {
if (!after(end_seq, TCP_SKB_CB(skb1)->end_seq)) {
/* All the bits are present. Drop. */
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOMERGE);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOMERGE);
tcp_drop(sk, skb);
skb = NULL;
tcp_dsack_set(sk, seq, end_seq);
@@ -4493,7 +4493,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
__skb_unlink(skb1, &tp->out_of_order_queue);
tcp_dsack_extend(sk, TCP_SKB_CB(skb1)->seq,
TCP_SKB_CB(skb1)->end_seq);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOMERGE);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFOMERGE);
tcp_drop(sk, skb1);
}
@@ -4658,7 +4658,7 @@ queue_and_out:
if (!after(TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt)) {
/* A retransmit, 2nd most common case. Force an immediate ack. */
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKLOST);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKLOST);
tcp_dsack_set(sk, TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq);
out_of_window:
@@ -4704,7 +4704,7 @@ static struct sk_buff *tcp_collapse_one(struct sock *sk, struct sk_buff *skb,
__skb_unlink(skb, list);
__kfree_skb(skb);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPRCVCOLLAPSED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPRCVCOLLAPSED);
return next;
}
@@ -4863,7 +4863,7 @@ static bool tcp_prune_ofo_queue(struct sock *sk)
bool res = false;
if (!skb_queue_empty(&tp->out_of_order_queue)) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_OFOPRUNED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_OFOPRUNED);
__skb_queue_purge(&tp->out_of_order_queue);
/* Reset SACK state. A conforming SACK implementation will
@@ -4892,7 +4892,7 @@ static int tcp_prune_queue(struct sock *sk)
SOCK_DEBUG(sk, "prune_queue: c=%x\n", tp->copied_seq);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_PRUNECALLED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_PRUNECALLED);
if (atomic_read(&sk->sk_rmem_alloc) >= sk->sk_rcvbuf)
tcp_clamp_window(sk);
@@ -4922,7 +4922,7 @@ static int tcp_prune_queue(struct sock *sk)
* drop receive data on the floor. It will get retransmitted
* and hopefully then we'll have sufficient space.
*/
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_RCVPRUNED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_RCVPRUNED);
/* Massive buffer overcommit. */
tp->pred_flags = 0;
@@ -5181,7 +5181,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb,
if (tcp_fast_parse_options(skb, th, tp) && tp->rx_opt.saw_tstamp &&
tcp_paws_discard(sk, skb)) {
if (!th->rst) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
if (!tcp_oow_rate_limited(sock_net(sk), skb,
LINUX_MIB_TCPACKSKIPPEDPAWS,
&tp->last_oow_ack_time))
@@ -5233,8 +5233,8 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb,
if (th->syn) {
syn_challenge:
if (syn_inerr)
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNCHALLENGE);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNCHALLENGE);
tcp_send_challenge_ack(sk, skb);
goto discard;
}
@@ -5349,7 +5349,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
tcp_data_snd_check(sk);
return;
} else { /* Header too small */
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
goto discard;
}
} else {
@@ -5377,7 +5377,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
__skb_pull(skb, tcp_header_len);
tcp_rcv_nxt_update(tp, TCP_SKB_CB(skb)->end_seq);
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPHPHITSTOUSER);
eaten = 1;
}
@@ -5400,7 +5400,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
tcp_rcv_rtt_measure_ts(sk, skb);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPHPHITS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPHPHITS);
/* Bulk data transfer: receiver */
eaten = tcp_queue_rcv(sk, skb, tcp_header_len,
@@ -5457,8 +5457,8 @@ step5:
return;
csum_error:
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
discard:
tcp_drop(sk, skb);
@@ -5550,13 +5550,13 @@ static bool tcp_rcv_fastopen_synack(struct sock *sk, struct sk_buff *synack,
break;
}
tcp_rearm_rto(sk);
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPFASTOPENACTIVEFAIL);
return true;
}
tp->syn_data_acked = tp->syn_data;
if (tp->syn_data_acked)
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPFASTOPENACTIVE);
tcp_fastopen_add_skb(sk, synack);
@@ -5592,7 +5592,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb,
if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr &&
!between(tp->rx_opt.rcv_tsecr, tp->retrans_stamp,
tcp_time_stamp)) {
- __NET_INC_STATS(sock_net(sk),
+ NET_INC_STATS(sock_net(sk),
LINUX_MIB_PAWSACTIVEREJECTED);
goto reset_and_undo;
}
@@ -5962,7 +5962,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
(TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq &&
after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt))) {
tcp_done(sk);
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA);
return 1;
}
@@ -6019,7 +6019,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
if (sk->sk_shutdown & RCV_SHUTDOWN) {
if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq &&
after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt)) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA);
tcp_reset(sk);
return 1;
}
@@ -6221,7 +6221,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
* timeout.
*/
if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
goto drop;
}
@@ -6268,7 +6268,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
if (dst && strict &&
!tcp_peer_is_proven(req, dst, true,
tmp_opt.saw_tstamp)) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSPASSIVEREJECTED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSPASSIVEREJECTED);
goto drop_and_release;
}
}
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 510f7a3c758b..fc89f2ce75f1 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -692,6 +692,7 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb)
offsetof(struct inet_timewait_sock, tw_bound_dev_if));
arg.tos = ip_hdr(skb)->tos;
+ preempt_disable();
ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
skb, &TCP_SKB_CB(skb)->header.h4.opt,
ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
@@ -699,6 +700,7 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb)
__TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
__TCP_INC_STATS(net, TCP_MIB_OUTRSTS);
+ preempt_enable();
#ifdef CONFIG_TCP_MD5SIG
out:
@@ -774,12 +776,14 @@ static void tcp_v4_send_ack(struct net *net,
if (oif)
arg.bound_dev_if = oif;
arg.tos = tos;
+ preempt_disable();
ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
skb, &TCP_SKB_CB(skb)->header.h4.opt,
ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
&arg, arg.iov[0].iov_len);
__TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
+ preempt_enable();
}
static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb)
@@ -1151,12 +1155,12 @@ static bool tcp_v4_inbound_md5_hash(const struct sock *sk,
return false;
if (hash_expected && !hash_location) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
return true;
}
if (!hash_expected && hash_location) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
return true;
}
@@ -1342,7 +1346,7 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb,
return newsk;
exit_overflow:
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
exit_nonewsk:
dst_release(dst);
exit:
@@ -1432,8 +1436,8 @@ discard:
return 0;
csum_err:
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
goto discard;
}
EXPORT_SYMBOL(tcp_v4_do_rcv);
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index ffbfecdae471..4b95ec4ed2c8 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -337,7 +337,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo)
* socket up. We've got bigger problems than
* non-graceful socket closings.
*/
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPTIMEWAITOVERFLOW);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPTIMEWAITOVERFLOW);
}
tcp_update_metrics(sk);
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index b1b2045ac3a9..411650ae3a19 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2206,14 +2206,13 @@ bool tcp_schedule_loss_probe(struct sock *sk)
/* Thanks to skb fast clones, we can detect if a prior transmit of
* a packet is still in a qdisc or driver queue.
* In this case, there is very little point doing a retransmit !
- * Note: This is called from BH context only.
*/
static bool skb_still_in_host_queue(const struct sock *sk,
const struct sk_buff *skb)
{
if (unlikely(skb_fclone_busy(sk, skb))) {
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES);
return true;
}
return false;
@@ -2275,7 +2274,7 @@ void tcp_send_loss_probe(struct sock *sk)
tp->tlp_high_seq = tp->snd_nxt;
probe_sent:
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSPROBES);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSPROBES);
/* Reset s.t. tcp_rearm_rto will restart timer from now */
inet_csk(sk)->icsk_pending = 0;
rearm_timer:
diff --git a/net/ipv4/tcp_recovery.c b/net/ipv4/tcp_recovery.c
index e0d0afaf15be..e36df4fcfeba 100644
--- a/net/ipv4/tcp_recovery.c
+++ b/net/ipv4/tcp_recovery.c
@@ -65,8 +65,8 @@ int tcp_rack_mark_lost(struct sock *sk)
if (scb->sacked & TCPCB_SACKED_RETRANS) {
scb->sacked &= ~TCPCB_SACKED_RETRANS;
tp->retrans_out -= tcp_skb_pcount(skb);
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPLOSTRETRANSMIT);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPLOSTRETRANSMIT);
}
} else if (!(scb->sacked & TCPCB_RETRANS)) {
/* Original data are sent sequentially so stop early
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 35f643d8ffbb..debdd8b33e69 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -162,8 +162,8 @@ static int tcp_write_timeout(struct sock *sk)
if (tp->syn_fastopen || tp->syn_data)
tcp_fastopen_cache_set(sk, 0, NULL, true, 0);
if (tp->syn_data && icsk->icsk_retransmits == 1)
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPFASTOPENACTIVEFAIL);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPFASTOPENACTIVEFAIL);
}
retry_until = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_syn_retries;
syn_set = true;
@@ -178,8 +178,8 @@ static int tcp_write_timeout(struct sock *sk)
tp->bytes_acked <= tp->rx_opt.mss_clamp) {
tcp_fastopen_cache_set(sk, 0, NULL, true, 0);
if (icsk->icsk_retransmits == net->ipv4.sysctl_tcp_retries1)
- __NET_INC_STATS(sock_net(sk),
- LINUX_MIB_TCPFASTOPENACTIVEFAIL);
+ NET_INC_STATS(sock_net(sk),
+ LINUX_MIB_TCPFASTOPENACTIVEFAIL);
}
/* Black hole detection */
tcp_mtu_probing(icsk, sk);
@@ -209,6 +209,7 @@ static int tcp_write_timeout(struct sock *sk)
return 0;
}
+/* Called with BH disabled */
void tcp_delack_timer_handler(struct sock *sk)
{
struct tcp_sock *tp = tcp_sk(sk);
@@ -493,6 +494,7 @@ out_reset_timer:
out:;
}
+/* Called with BH disabled */
void tcp_write_timer_handler(struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 52914714b923..7bdc9c9c231b 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -649,12 +649,12 @@ static bool tcp_v6_inbound_md5_hash(const struct sock *sk,
return false;
if (hash_expected && !hash_location) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
return true;
}
if (!hash_expected && hash_location) {
- __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
+ NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
return true;
}
@@ -825,9 +825,9 @@ static void tcp_v6_send_response(const struct sock *sk, struct sk_buff *skb, u32
if (!IS_ERR(dst)) {
skb_dst_set(buff, dst);
ip6_xmit(ctl_sk, buff, &fl6, NULL, tclass);
- __TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
+ TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
if (rst)
- __TCP_INC_STATS(net, TCP_MIB_OUTRSTS);
+ TCP_INC_STATS(net, TCP_MIB_OUTRSTS);
return;
}
@@ -1276,8 +1276,8 @@ discard:
kfree_skb(skb);
return 0;
csum_err:
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
- __TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
+ TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
goto discard;
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH net-next 4/6] udp: prepare for non BH masking at backlog processing
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
UDP uses the generic socket backlog code, and this will soon
be changed to not disable BH when protocol is called back.
We need to use appropriate SNMP accessors.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/ipv4/udp.c | 4 ++--
net/ipv6/udp.c | 4 ++--
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 093284c5c03b..f67f52ba4809 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1514,9 +1514,9 @@ static int __udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
/* Note that an ENOMEM error is charged twice */
if (rc == -ENOMEM)
- __UDP_INC_STATS(sock_net(sk), UDP_MIB_RCVBUFERRORS,
+ UDP_INC_STATS(sock_net(sk), UDP_MIB_RCVBUFERRORS,
is_udplite);
- __UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
+ UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
kfree_skb(skb);
trace_udp_fail_queue_rcv_skb(rc, sk);
return -1;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 1ba5a74ac18f..f911c63f79e6 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -570,9 +570,9 @@ static int __udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
/* Note that an ENOMEM error is charged twice */
if (rc == -ENOMEM)
- __UDP6_INC_STATS(sock_net(sk),
+ UDP6_INC_STATS(sock_net(sk),
UDP_MIB_RCVBUFERRORS, is_udplite);
- __UDP6_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
+ UDP6_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
kfree_skb(skb);
return -1;
}
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH net-next 5/6] sctp: prepare for socket backlog behavior change
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
sctp_inq_push() will soon be called without BH being blocked
when generic socket code flushes the socket backlog.
It is very possible SCTP can be converted to not rely on BH,
but this needs to be done by SCTP experts.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/sctp/inqueue.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/net/sctp/inqueue.c b/net/sctp/inqueue.c
index b335ffcef0b9..9d87bba0ff1d 100644
--- a/net/sctp/inqueue.c
+++ b/net/sctp/inqueue.c
@@ -89,10 +89,12 @@ void sctp_inq_push(struct sctp_inq *q, struct sctp_chunk *chunk)
* Eventually, we should clean up inqueue to not rely
* on the BH related data structures.
*/
+ local_bh_disable();
list_add_tail(&chunk->list, &q->in_chunk_list);
if (chunk->asoc)
chunk->asoc->stats.ipackets++;
q->immediate.func(&q->immediate);
+ local_bh_enable();
}
/* Peek at the next chunk on the inqeue. */
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* [PATCH net-next 6/6] net: do not block BH while processing socket backlog
From: Eric Dumazet @ 2016-04-28 5:25 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev, Eric Dumazet, Eric Dumazet
In-Reply-To: <1461821152-23200-1-git-send-email-edumazet@google.com>
Socket backlog processing is a major latency source.
With current TCP socket sk_rcvbuf limits, I have sampled __release_sock()
holding cpu for more than 5 ms, and packets being dropped by the NIC
once ring buffer is filled.
All users are now ready to be called from process context,
we can unblock BH and let interrupts be serviced faster.
cond_resched_softirq() could be removed, as it has no more user.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
net/core/sock.c | 22 ++++++++--------------
1 file changed, 8 insertions(+), 14 deletions(-)
diff --git a/net/core/sock.c b/net/core/sock.c
index e16a5db853c6..70744dbb6c3f 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2019,33 +2019,27 @@ static void __release_sock(struct sock *sk)
__releases(&sk->sk_lock.slock)
__acquires(&sk->sk_lock.slock)
{
- struct sk_buff *skb = sk->sk_backlog.head;
+ struct sk_buff *skb, *next;
- do {
+ while ((skb = sk->sk_backlog.head) != NULL) {
sk->sk_backlog.head = sk->sk_backlog.tail = NULL;
- bh_unlock_sock(sk);
- do {
- struct sk_buff *next = skb->next;
+ spin_unlock_bh(&sk->sk_lock.slock);
+ do {
+ next = skb->next;
prefetch(next);
WARN_ON_ONCE(skb_dst_is_noref(skb));
skb->next = NULL;
sk_backlog_rcv(sk, skb);
- /*
- * We are in process context here with softirqs
- * disabled, use cond_resched_softirq() to preempt.
- * This is safe to do because we've taken the backlog
- * queue private:
- */
- cond_resched_softirq();
+ cond_resched();
skb = next;
} while (skb != NULL);
- bh_lock_sock(sk);
- } while ((skb = sk->sk_backlog.head) != NULL);
+ spin_lock_bh(&sk->sk_lock.slock);
+ }
/*
* Doing the zeroing here guarantee we can not loop forever
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* Re: [PATCH] vhost_net: stop polling socket during rx processing
From: Jason Wang @ 2016-04-28 6:19 UTC (permalink / raw)
To: Michael S. Tsirkin; +Cc: netdev, linux-kernel, kvm, virtualization
In-Reply-To: <20160427141317-mutt-send-email-mst@redhat.com>
On 04/27/2016 07:28 PM, Michael S. Tsirkin wrote:
> On Tue, Apr 26, 2016 at 03:35:53AM -0400, Jason Wang wrote:
>> We don't stop polling socket during rx processing, this will lead
>> unnecessary wakeups from under layer net devices (E.g
>> sock_def_readable() form tun). Rx will be slowed down in this
>> way. This patch avoids this by stop polling socket during rx
>> processing. A small drawback is that this introduces some overheads in
>> light load case because of the extra start/stop polling, but single
>> netperf TCP_RR does not notice any change. In a super heavy load case,
>> e.g using pktgen to inject packet to guest, we get about ~17%
>> improvement on pps:
>>
>> before: ~1370000 pkt/s
>> after: ~1500000 pkt/s
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
> Acked-by: Michael S. Tsirkin <mst@redhat.com>
>
> There is one other possible enhancement: we actually have the wait queue
> lock taken in _wake_up, but we give it up only to take it again in the
> handler.
>
> It would be nicer to just remove the entry when we wake
> the vhost thread. Re-add it if required.
> I think that something like the below would give you the necessary API.
> Pls feel free to use it if you are going to implement a patch on top
> doing this - that's not a reason not to include this simple patch
> though.
Thanks, this looks useful, will give it a try.
>
> --->
>
> wait: add API to drop a wait_queue_t entry from wake up handler
>
> A wake up handler might want to remove its own wait queue entry to avoid
> future wakeups. In particular, vhost has such a need. As wait queue
> lock is already taken, all we need is an API to remove the entry without
> wait_queue_head_t which isn't currently accessible to wake up handlers.
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
>
> ---
>
> diff --git a/include/linux/wait.h b/include/linux/wait.h
> index 27d7a0a..9c6604b 100644
> --- a/include/linux/wait.h
> +++ b/include/linux/wait.h
> @@ -191,11 +191,17 @@ __add_wait_queue_tail_exclusive(wait_queue_head_t *q, wait_queue_t *wait)
> }
>
> static inline void
> -__remove_wait_queue(wait_queue_head_t *head, wait_queue_t *old)
> +__remove_wait_queue_entry(wait_queue_t *old)
> {
> list_del(&old->task_list);
> }
>
> +static inline void
> +__remove_wait_queue(wait_queue_head_t *head, wait_queue_t *old)
> +{
> + __remove_wait_queue_entry(old);
> +}
> +
> typedef int wait_bit_action_f(struct wait_bit_key *, int mode);
> void __wake_up(wait_queue_head_t *q, unsigned int mode, int nr, void *key);
> void __wake_up_locked_key(wait_queue_head_t *q, unsigned int mode, void *key);
^ permalink raw reply
* Re: [RFC PATCH V2 1/2] vhost: convert pre sorted vhost memory array to interval tree
From: Jason Wang @ 2016-04-28 6:20 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: kvm, qemu-devel, netdev, linux-kernel, peterx, virtualization,
pbonzini
In-Reply-To: <20160427142948-mutt-send-email-mst@redhat.com>
On 04/27/2016 07:30 PM, Michael S. Tsirkin wrote:
> On Fri, Mar 25, 2016 at 10:34:33AM +0800, Jason Wang wrote:
>> > Current pre-sorted memory region array has some limitations for future
>> > device IOTLB conversion:
>> >
>> > 1) need extra work for adding and removing a single region, and it's
>> > expected to be slow because of sorting or memory re-allocation.
>> > 2) need extra work of removing a large range which may intersect
>> > several regions with different size.
>> > 3) need trick for a replacement policy like LRU
>> >
>> > To overcome the above shortcomings, this patch convert it to interval
>> > tree which can easily address the above issue with almost no extra
>> > work.
>> >
>> > The patch could be used for:
>> >
>> > - Extend the current API and only let the userspace to send diffs of
>> > memory table.
>> > - Simplify Device IOTLB implementation.
> Does this affect performance at all?
>
In pktgen test, no difference.
Thanks
^ permalink raw reply
* Re: [PATCH v2] can: rcar_canfd: Add Renesas R-Car CAN FD driver
From: Oliver Hartkopp @ 2016-04-28 6:27 UTC (permalink / raw)
To: Ramesh Shanmugasundaram, wg@grandegger.com, robh+dt@kernel.org,
pawel.moll@arm.com, mark.rutland@arm.com,
ijc+devicetree@hellion.org.uk, galak@codeaurora.org,
corbet@lwn.net, mkl@pengutronix.de
Cc: linux-renesas-soc@vger.kernel.org, devicetree@vger.kernel.org,
linux-can@vger.kernel.org, netdev@vger.kernel.org,
linux-doc@vger.kernel.org, geert+renesas@glider.be,
Chris Paterson
In-Reply-To: <KL1PR06MB10312B266592335D965823AFC3960@KL1PR06MB1031.apcprd06.prod.outlook.com>
Hello Ramesh,
please send out a new v3 patchset to trigger the process again :-)
Best regards,
Oliver
On 04/13/2016 08:25 AM, Ramesh Shanmugasundaram wrote:
> HI Marc,
>
> Gentle reminder!
> Are you happy with the open comment's disposition? I can send a next version of patch if we have a closure on current set of comments.
>
> Thanks,
> Ramesh
>
>> -----Original Message-----
>> From: Ramesh Shanmugasundaram
>> Sent: 01 April 2016 13:49
>> To: Ramesh Shanmugasundaram <ramesh.shanmugasundaram@bp.renesas.com>;
>> wg@grandegger.com; robh+dt@kernel.org; pawel.moll@arm.com;
>> mark.rutland@arm.com; ijc+devicetree@hellion.org.uk; galak@codeaurora.org;
>> corbet@lwn.net
>> Cc: linux-renesas-soc@vger.kernel.org; devicetree@vger.kernel.org; linux-
>> can@vger.kernel.org; netdev@vger.kernel.org; linux-doc@vger.kernel.org;
>> geert+renesas@glider.be; Chris Paterson <Chris.Paterson2@renesas.com>
>> Subject: RE: [PATCH v2] can: rcar_canfd: Add Renesas R-Car CAN FD driver
>>
>> Hi Marc,
>>
>> Thanks for your time & review comments.
>>
>>> -----Original Message-----
>> (...)
>>>> +#define RCANFD_NAPI_WEIGHT 8 /* Rx poll quota */
>>>> +
>>>> +#define RCANFD_NUM_CHANNELS 2
>>>> +#define RCANFD_CHANNELS_MASK 0x3 /* Two channels max */
>>>
>>> (BIT(RCANFD_NUM_CHANNELS) - 1
>>
>> OK
>>
>>>
>>>> +
>>>> +/* Rx FIFO is a global resource of the controller. There are 8 such
>>> FIFOs
>>>> + * available. Each channel gets a dedicated Rx FIFO (i.e.) the
>>>> + channel
>> (...)
>>>> +#define RCANFD_CMFIFO_CFDLC(x) (((x) & 0xf) << 28)
>>>> +#define RCANFD_CMFIFO_CFPTR(x) (((x) & 0xfff) << 16)
>>>> +#define RCANFD_CMFIFO_CFTS(x) (((x) & 0xff) << 0)
>>>> +
>>>> +/* Global Test Config register */
>>>> +#define RCANFD_GTSTCFG_C0CBCE BIT(0)
>>>> +#define RCANFD_GTSTCFG_C1CBCE BIT(1)
>>>> +
>>>> +#define RCANFD_GTSTCTR_ICBCTME BIT(0)
>>>> +
>>>> +/* AFL Rx rules registers */
>>>> +#define RCANFD_AFLCFG_SETRNC0(x) (((x) & 0xff) << 24)
>>>> +#define RCANFD_AFLCFG_SETRNC1(x) (((x) & 0xff) << 16)
>>>
>>> What about something like:
>>>
>>> #define RCANFD_AFLCFG_SETRNC(n, x) (((x) & 0xff) << (24 - n * 8))
>>>
>>> This will save some if()s in the code
>>
>> Nice :-). Will update.
>>
>>>
>>>> +#define RCANFD_AFLCFG_GETRNC0(x) (((x) >> 24) & 0xff)
>>>> +#define RCANFD_AFLCFG_GETRNC1(x) (((x) >> 16) & 0xff)
>>>> +
>>>> +#define RCANFD_AFL_PAGENUM(entry) ((entry) / 16)
>> (...)
>>>> +#define rcar_canfd_read(priv, offset) \
>>>> + readl(priv->base + (offset))
>>>> +#define rcar_canfd_write(priv, offset, val) \
>>>> + writel(val, priv->base + (offset))
>>>> +#define rcar_canfd_set_bit(priv, reg, val) \
>>>> + rcar_canfd_update(val, val, priv->base + (reg))
>>>> +#define rcar_canfd_clear_bit(priv, reg, val) \
>>>> + rcar_canfd_update(val, 0, priv->base + (reg))
>>>> +#define rcar_canfd_update_bit(priv, reg, mask, val) \
>>>> + rcar_canfd_update(mask, val, priv->base + (reg))
>>>
>>> please use static inline functions instead of defines.
>>
>> OK.
>>
>>>
>>>> +
>>>> +static void rcar_canfd_get_data(struct canfd_frame *cf,
>>>> + struct rcar_canfd_channel *priv, u32 off)
>>>
>>> Please use "struct rcar_canfd_channel *priv" as first argument, struct
>>> canfd_frame *cf as second. Remove off, as the offset is already
>>> defined by the channel.
>>
>> I'll re-order priv, cf as you mentioned. I'll leave "off" arg as it is
>> because it is based on FIFO number of channel + mode (CAN only or CANFD
>> only). Otherwise, I will have to add another check inside this function
>> for mode.
>>
>>>> +{
>>>> + u32 i, lwords;
>>>> +
>>>> + lwords = cf->len / sizeof(u32);
>>>> + if (cf->len % sizeof(u32))
>>>> + lwords++;
>>>
>>> Use DIV_ROUND_UP() instead of open coding it.
>>
>> Agreed. Thanks.
>>
>>>
>>>> + for (i = 0; i < lwords; i++)
>>>> + *((u32 *)cf->data + i) =
>>>> + rcar_canfd_read(priv, off + (i * sizeof(u32))); }
>>>> +
>>>> +static void rcar_canfd_put_data(struct canfd_frame *cf,
>>>> + struct rcar_canfd_channel *priv, u32 off)
>>>
>>> same here
>>
>> Yes (same as _get_data)
>>
>>>
>>>> +{
>>>> + u32 i, j, lwords, leftover;
>>>> + u32 data = 0;
>>>> +
>>>> + lwords = cf->len / sizeof(u32);
>>>> + leftover = cf->len % sizeof(u32);
>>>> + for (i = 0; i < lwords; i++)
>>>> + rcar_canfd_write(priv, off + (i * sizeof(u32)),
>>>> + *((u32 *)cf->data + i));
>>>
>>> Here you don't convert the endianess...
>>
>> Yes
>>
>>>> +
>>>> + if (leftover) {
>>>> + u8 *p = (u8 *)((u32 *)cf->data + i);
>>>> +
>>>> + for (j = 0; j < leftover; j++)
>>>> + data |= p[j] << (j * 8);
>>>
>>> ...here you do an implicit endianess conversion. "data" is little
>>> endian, while p[j] is big endian.
>>
>> Not sure I got the question correctly.
>> Controller expectation of data bytes in 32bit register is bits[7:0] =
>> byte0, bits[15:8] = byte1 and so on - little endian.
>> For e.g. if cf->data points to byte stream H'112233445566 (cf->data[0] =
>> 0x11), first rcar_canfd_write will write 0x44332211 value to register. Yes
>> the host cpu is assumed little endian. In leftover case, data will be
>> 0x00006655 - again little endian.
>> I think I should remove this leftover logic and just mask the unused bits
>> to zero as cf->data is pre-allocated for max length anyway.
>>
>> p[j] is a byte?? Am I missing something?
>>
>>>
>>>> + rcar_canfd_write(priv, off + (i * sizeof(u32)), data);
>>>> + }
>>>
>>> Have you tested to send CAN frames with len != n*4 against a different
>>> controller?
>>
>> Yes, with Vector VN1630A.
>>
>>>> +}
>>>> +
>>>> +static void rcar_canfd_tx_failure_cleanup(struct net_device *ndev)
>>>> +{
>>>> + u32 i;
>>>> +
>>>> + for (i = 0; i < RCANFD_FIFO_DEPTH; i++)
>>>> + can_free_echo_skb(ndev, i);
>>>> +}
>>>> +
>> (...)
>>>> +static void rcar_canfd_tx_done(struct net_device *ndev) {
>>>> + struct rcar_canfd_channel *priv = netdev_priv(ndev);
>>>> + struct net_device_stats *stats = &ndev->stats;
>>>> + u32 sts;
>>>> + unsigned long flags;
>>>> + u32 ch = priv->channel;
>>>> +
>>>> + do {
>>>
>>> You should iterare over all pending CAN frames:
>>>
>>>> for (/* nix */; (priv->tx_head - priv->tx_tail) > 0; priv-
>>>> tx_tail++) {
>>>
>>
>> Yes, current code does iterate over pending tx'ed frames. If we use this
>> for loop semantics, we may have to protect the whole loop with no real
>> benefit.
>>
>>>
>>>> + u8 unsent, sent;
>>>> +
>>>> + sent = priv->tx_tail % RCANFD_FIFO_DEPTH;
>>>
>>> and check here, if that packet has really been tramsitted. Exit the
>>> loop otherweise.
>>
>> We are here because of tx_done and hence no need to check tx done again
>> for the first iteration. Hence the do-while loop. Checks further down
>> takes care of the need for more iterations/pending tx'ed frames.
>>
>>>
>>>> + stats->tx_packets++;
>>>> + stats->tx_bytes += priv->tx_len[sent];
>>>> + priv->tx_len[sent] = 0;
>>>> + can_get_echo_skb(ndev, sent);
>>>> +
>>>> + spin_lock_irqsave(&priv->tx_lock, flags);
>>>
>>> What does the tx_lock protect? The tx path is per channel, isn't it?
>>
>> You are right - tx path is per channel. In tx path, head & tail are also
>> used to determine when to stop/wakeup netif queue. They are incremented &
>> compared in different contexts to make this decision. Per channel tx_lock
>> protects this critical section.
>>
>>>
>>>> + priv->tx_tail++;
>>>> + sts = rcar_canfd_read(priv,
>>>> + RCANFD_CFSTS(ch, RCANFD_CFFIFO_IDX));
>>>> + unsent = RCANFD_CMFIFO_CFMC(sts);
>>>> +
>>>> + /* Wake producer only when there is room */
>>>> + if (unsent != RCANFD_FIFO_DEPTH)
>>>> + netif_wake_queue(ndev);
>>>
>>> Move the netif_wake_queue() out of the loop.
>>
>> With the tx logic mentioned above, I think keeping it in the loop seems
>> better. For e.g. say cpu1 pumps 8 frames from an app loop and cpu0 handles
>> tx_done interrupt handling: cpu1 would have stopped the queue because FIFO
>> is full -> cpu0 gets tx_done interrupt and by the time it checks "unsent"
>> there could be one or more frames transmitted by device (i.e.) there would
>> be more space in fifo. It is better to wakeup the netif queue then and
>> there so that app running on cpu1 can pump more. If we move it out of the
>> loop we wake up the queue only in the end. Have I missed anything?
>>
>>>
>>>> +
>>>> + if (priv->tx_head - priv->tx_tail <= unsent) {
>>>> + spin_unlock_irqrestore(&priv->tx_lock, flags);
>>>> + break;
>>>> + }
>>>> + spin_unlock_irqrestore(&priv->tx_lock, flags);
>>>> +
>>>> + } while (1);
>>>> +
>>>> + /* Clear interrupt */
>>>> + rcar_canfd_write(priv, RCANFD_CFSTS(ch, RCANFD_CFFIFO_IDX),
>>>> + sts & ~RCANFD_CMFIFO_CFTXIF);
>>>> + can_led_event(ndev, CAN_LED_EVENT_TX); }
>>>> +
>>>> + if (cf->can_id & CAN_RTR_FLAG)
>>>> + id |= RCANFD_CMFIFO_CFRTR;
>>>> +
>>>> + rcar_canfd_write(priv, RCANFD_F_CFID(ch, RCANFD_CFFIFO_IDX),
>>>> + id);
>>>> + ptr = RCANFD_CMFIFO_CFDLC(can_len2dlc(cf->len));
>>>
>>> ptr usually means pointer, better call it dlc.
>>
>> I used the register name "ptr". OK, will change it do dlc.
>>
>>>
>>>> + rcar_canfd_write(priv, RCANFD_F_CFPTR(ch, RCANFD_CFFIFO_IDX),
>>>> + ptr);
>>>> +
>> (...)
>>>> + can_put_echo_skb(skb, ndev, priv->tx_head % RCANFD_FIFO_DEPTH);
>>>> +
>>>> + spin_lock_irqsave(&priv->tx_lock, flags);
>>>> + priv->tx_head++;
>>>> +
>>>> + /* Start Tx: Write 0xff to CFPC to increment the CPU-side
>>>> + * pointer for the Common FIFO
>>>> + */
>>>> + rcar_canfd_write(priv, RCANFD_CFPCTR(ch, RCANFD_CFFIFO_IDX),
>>>> +0xff);
>>>> +
>>>> + /* Stop the queue if we've filled all FIFO entries */
>>>> + if (priv->tx_head - priv->tx_tail >= RCANFD_FIFO_DEPTH)
>>>> + netif_stop_queue(ndev);
>>>
>>> Please move the check of stop_queue, before starting the send.
>>
>> OK.
>>
>>>
>>>> +
>>>> + spin_unlock_irqrestore(&priv->tx_lock, flags);
>>>> + return NETDEV_TX_OK;
>>>> +}
>>>> +
>> (...)
>>>> +{
>>>> + struct rcar_canfd_channel *priv =
>>>> + container_of(napi, struct rcar_canfd_channel, napi);
>>>> + int num_pkts;
>>>> + u32 sts;
>>>> + u32 ch = priv->channel;
>>>> + u32 ridx = ch + RCANFD_RFFIFO_IDX;
>>>> +
>>>> + for (num_pkts = 0; num_pkts < quota; num_pkts++) {
>>>> + sts = rcar_canfd_read(priv, RCANFD_RFSTS(ridx));
>>>> + /* Clear interrupt bit */
>>>> + if (sts & RCANFD_RFFIFO_RFIF)
>>>> + rcar_canfd_write(priv, RCANFD_RFSTS(ridx),
>>>> + sts & ~RCANFD_RFFIFO_RFIF);
>>>> +
>>>> + /* Check FIFO empty condition */
>>>> + if (sts & RCANFD_RFFIFO_RFEMP)
>>>> + break;
>>>> +
>>>> + rcar_canfd_rx_pkt(priv);
>>>
>>> This sequence looks strange. You first conditionally ack the interrupt
>>> then you check for empty fifo, then read the CAN frame. I would assume
>>> that you first check if there's a CAN frame, read it and then clear
>>> the interrupt.
>>
>> Yes, I shall re-arrange the sequence as you mentioned.
>>
>>>
>>>> + }
>>>> +
>>>> + /* All packets processed */
>>>> + if (num_pkts < quota) {
>>>> + napi_complete(napi);
>>>> + /* Enable Rx FIFO interrupts */
>>>> + rcar_canfd_set_bit(priv, RCANFD_RFCC(ridx),
>>> RCANFD_RFFIFO_RFIE);
>> (...)
>>>> + for_each_set_bit(ch, &gpriv->channels_mask, RCANFD_NUM_CHANNELS) {
>>>> + err = rcar_canfd_channel_probe(gpriv, ch);
>>>> + if (err)
>>>> + goto fail_channel;
>>>> + }
>>>
>>> Should the CAN IP core be clocked the whole time? What about shuting
>>> down the clock and enabling it on ifup?
>>
>> The fCAN clock is enabled only on ifup of one of the channels. However,
>> the peripheral clock is enabled during probe to bring the controller to
>> Global Operational mode. This clock cannot be turned off with the register
>> values & mode retained.
>>
>>>> +
>>>> + platform_set_drvdata(pdev, gpriv);
>>>> + dev_info(&pdev->dev, "global operational state (clk %d)\n",
>>>> + gpriv->clock_select);
>> (...)
>>>> + rcar_canfd_reset_controller(gpriv);
>>>> + rcar_canfd_disable_global_interrupts(gpriv);
>>>> +
>>>> + for_each_set_bit(ch, &gpriv->channels_mask, RCANFD_NUM_CHANNELS) {
>>>> + priv = gpriv->ch[ch];
>>>> + if (priv) {
>>>
>>> This should always be true.
>>
>> I agree. I thought I cleaned this up but it's in my local commits :-(
>>
>>>
>>>> + rcar_canfd_disable_channel_interrupts(priv);
>>>> + unregister_candev(priv->ndev);
>>>> + netif_napi_del(&priv->napi);
>>>> + free_candev(priv->ndev);
>>>
>>> Please make use of rcar_canfd_channel_remove(), as you already have
>>> the function.
>>
>> Yes.
>>
>> Thanks,
>> Ramesh
^ permalink raw reply
* Re: [RFC PATCH V2 2/2] vhost: device IOTLB API
From: Jason Wang @ 2016-04-28 6:37 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: kvm, qemu-devel, netdev, linux-kernel, peterx, virtualization,
pbonzini
In-Reply-To: <20160427143057-mutt-send-email-mst@redhat.com>
On 04/27/2016 07:45 PM, Michael S. Tsirkin wrote:
> On Fri, Mar 25, 2016 at 10:34:34AM +0800, Jason Wang wrote:
>> This patch tries to implement an device IOTLB for vhost. This could be
>> used with for co-operation with userspace(qemu) implementation of DMA
>> remapping.
>>
>> The idea is simple. When vhost meets an IOTLB miss, it will request
>> the assistance of userspace to do the translation, this is done
>> through:
>>
>> - Fill the translation request in a preset userspace address (This
>> address is set through ioctl VHOST_SET_IOTLB_REQUEST_ENTRY).
>> - Notify userspace through eventfd (This eventfd was set through ioctl
>> VHOST_SET_IOTLB_FD).
> Why use an eventfd for this?
The aim is to implement the API all through ioctls.
> We use them for interrupts because
> that happens to be what kvm wants, but here - why don't we
> just add a generic support for reading out events
> on the vhost fd itself?
I've considered this approach, but what's the advantages of this? I mean
looks like all other ioctls could be done through vhost fd
reading/writing too.
>
>> - device IOTLB were started and stopped through VHOST_RUN_IOTLB ioctl
>>
>> When userspace finishes the translation, it will update the vhost
>> IOTLB through VHOST_UPDATE_IOTLB ioctl. Userspace is also in charge of
>> snooping the IOTLB invalidation of IOMMU IOTLB and use
>> VHOST_UPDATE_IOTLB to invalidate the possible entry in vhost.
> There's one problem here, and that is that VQs still do not undergo
> translation. In theory VQ could be mapped in such a way
> that it's not contigious in userspace memory.
I'm not sure I get the issue, current vhost API support setting
desc_user_addr, used_user_addr and avail_user_addr independently. So
looks ok? If not, looks not a problem to device IOTLB API itself.
>
>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
> What limits amount of entries that kernel keeps around?
It depends on guest working set I think. Looking at
http://dpdk.org/doc/guides/linux_gsg/sys_reqs.html:
- For 2MB page size in guest, it suggests hugepages=1024
- For 1GB page size, it suggests a hugepages=4
So I choose 2048 to make sure it can cover this.
> Do we want at least a mod parameter for this?
Maybe.
>
>> ---
>> drivers/vhost/net.c | 6 +-
>> drivers/vhost/vhost.c | 301 +++++++++++++++++++++++++++++++++++++++------
>> drivers/vhost/vhost.h | 17 ++-
>> fs/eventfd.c | 3 +-
>> include/uapi/linux/vhost.h | 35 ++++++
>> 5 files changed, 320 insertions(+), 42 deletions(-)
>>
[...]
>> +struct vhost_iotlb_entry {
>> + __u64 iova;
>> + __u64 size;
>> + __u64 userspace_addr;
> Alignment requirements?
The API does not require any alignment. Will add a comment for this.
>
>> + struct {
>> +#define VHOST_ACCESS_RO 0x1
>> +#define VHOST_ACCESS_WO 0x2
>> +#define VHOST_ACCESS_RW 0x3
>> + __u8 perm;
>> +#define VHOST_IOTLB_MISS 1
>> +#define VHOST_IOTLB_UPDATE 2
>> +#define VHOST_IOTLB_INVALIDATE 3
>> + __u8 type;
>> +#define VHOST_IOTLB_INVALID 0x1
>> +#define VHOST_IOTLB_VALID 0x2
>> + __u8 valid;
> why do we need this flag?
Useless, will remove.
>
>> + __u8 u8_padding;
>> + __u32 padding;
>> + } flags;
>> +};
>> +
>> +struct vhost_vring_iotlb_entry {
>> + unsigned int index;
>> + __u64 userspace_addr;
>> +};
>> +
>> struct vhost_memory_region {
>> __u64 guest_phys_addr;
>> __u64 memory_size; /* bytes */
>> @@ -127,6 +153,15 @@ struct vhost_memory {
>> /* Set eventfd to signal an error */
>> #define VHOST_SET_VRING_ERR _IOW(VHOST_VIRTIO, 0x22, struct vhost_vring_file)
>>
>> +/* IOTLB */
>> +/* Specify an eventfd file descriptor to signle on IOTLB miss */
> typo
Will fix it.
>
>> +#define VHOST_SET_VRING_IOTLB_CALL _IOW(VHOST_VIRTIO, 0x23, struct \
>> + vhost_vring_file)
>> +#define VHOST_SET_VRING_IOTLB_REQUEST _IOW(VHOST_VIRTIO, 0x25, struct \
>> + vhost_vring_iotlb_entry)
>> +#define VHOST_UPDATE_IOTLB _IOW(VHOST_VIRTIO, 0x24, struct vhost_iotlb_entry)
>> +#define VHOST_RUN_IOTLB _IOW(VHOST_VIRTIO, 0x26, int)
>> +
> Is the assumption that userspace must dedicate a thread to running the iotlb?
> I dislike this one.
> Please support asynchronous APIs at least optionally to make
> userspace make its own threading decisions.
Nope, my qemu patches does not use a dedicated thread. This API is used
to start or top DMAR according to e.g whether guest enable DMAR for
intel IOMMU.
>
>> /* VHOST_NET specific defines */
>>
>> /* Attach virtio net ring to a raw socket, or tap device.
> Don't we need a feature bit for this?
Yes we need it. The feature bit is not considered in this patch and
looks like it was still under discussion. After we finalize it, I will add.
> Are we short on feature bits? If yes maybe it's time to add
> something like PROTOCOL_FEATURES that we have in vhost-user.
>
I believe it can just work like VERSION_1, or is there anything I missed?
^ permalink raw reply
* [PATCH net-next 00/10] net: hns: add support of debug dsaf device
From: Yisen Zhuang @ 2016-04-28 7:01 UTC (permalink / raw)
To: devicetree, netdev, linux-arm-kernel
Cc: robh+dt, pawel.moll, mark.rutland, ijc+devicetree, galak, davem,
will.deacon, catalin.marinas, yankejian, huangdaode, salil.mehta,
lipeng321, liguozhu, xieqianqian, xuwei5, linuxarm
There are two kinds of dsaf device in hns, one is for service ports,
contains crossbar in it, can work under different mode. Another is for
debug port, only can work under single port mode. The current code only
declared a dsaf device for both service ports and debug ports.It is not so
readability. This patch separate it to three platform devices to make the
code simple and more readability.
all port in one platform device(old):
CPU
|
| DSAF(one platform device)
-------------------------------------------------------------- /
| | | | | /
| PPE PPE PPE | /
| | | | | /
| | | | | /
| crossbar | | | /
| | | | |/
| ----------------------------------- | | |
| | | | | | | | | |
| | | | | | | | | |
| MAC MAC MAC MAC MAC MAC MAC MAC |
| | | | | | | | | |
--------------------------------------------------------------
| | | | | | | |
PHY PHY PHY PHY PHY PHY PHY PHY
separate all ports to three platform(new):
CPU
|
-----------------------------------
| | |
---------------------------------------------- --------- ---------
| | | | | | | |
| PPE | | PPE | | PPE |
| | | | | | | | |
| | | | | | | | |
| crossbar | | | | | | |
| | | | | | | | |
| ---------------------------------- | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| MAC MAC MAC MAC MAC MAC | | MAC | | MAC |
| | | | | | | | | | | | | |
---------------------------------------------- --------- ---------
| | | | | | \ / | / |
PHY PHY PHY PHY PHY PHY \ / PHY / PHY
\ / /
\ / /
DSAF(three platform device)
I have tested this patchset on d02 board, it is compatible with the dts in
d02 board.
For more details, please see individual patches.
MBR.
Daode Huang (3):
net: hns: set debug port irq index to 0
net: hns: add syscon operation for dsaf
net: hns: sort the header file by alphabetical order
Yisen Zhuang (6):
net: hns: add attribute port-idx-in-ae in enet node.
net: hns: add attribute reset-field-offset for dsaf node
net: hns: separate debug dsaf device from service dsaf device
net: hns: add attribute cpld_ctrl for dsaf port node
net: hns: add attribute port-rst-offset for dsaf port node
net: hns: add attribute port-mode-offset for dsaf port node
huangdaode (1):
net: hns: add a new dsaf mode for debug port
.../devicetree/bindings/net/hisilicon-hns-dsaf.txt | 60 +++++-
.../devicetree/bindings/net/hisilicon-hns-nic.txt | 30 ++-
drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c | 73 ++-----
drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c | 231 ++++++++++++++-------
drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h | 12 +-
drivers/net/ethernet/hisilicon/hns/hns_dsaf_main.c | 141 ++++++++-----
drivers/net/ethernet/hisilicon/hns/hns_dsaf_main.h | 16 +-
drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c | 173 +++++++++------
drivers/net/ethernet/hisilicon/hns/hns_dsaf_ppe.c | 51 +----
drivers/net/ethernet/hisilicon/hns/hns_dsaf_ppe.h | 1 -
drivers/net/ethernet/hisilicon/hns/hns_dsaf_rcb.c | 187 +++++++----------
drivers/net/ethernet/hisilicon/hns/hns_dsaf_rcb.h | 2 +-
drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h | 38 ++--
drivers/net/ethernet/hisilicon/hns/hns_enet.c | 17 +-
drivers/net/ethernet/hisilicon/hns/hns_enet.h | 3 +
15 files changed, 584 insertions(+), 451 deletions(-)
--
1.9.1
^ permalink raw reply
* [PATCH net-next 09/10] net: hns: add attribute port-rst-offset for dsaf port node
From: Yisen Zhuang @ 2016-04-28 7:01 UTC (permalink / raw)
To: devicetree, netdev, linux-arm-kernel
Cc: robh+dt, pawel.moll, mark.rutland, ijc+devicetree, galak, davem,
will.deacon, catalin.marinas, yankejian, huangdaode, salil.mehta,
lipeng321, liguozhu, xieqianqian, xuwei5, linuxarm
In-Reply-To: <1461826902-51368-1-git-send-email-Yisen.Zhuang@huawei.com>
The reset offset for each port in a dsaf is different. The current code is
not so readability. This patch adds configuration named port-rst-offset to
make the code simple and more readability. If this attribute isn't exists,
default value of this attribute is equal to its port index.
Signed-off-by: Yisen Zhuang <yisen.zhuang@huawei.com>
---
.../devicetree/bindings/net/hisilicon-hns-dsaf.txt | 4 +++-
drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c | 10 +++++++++
drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h | 1 +
drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c | 26 +++++++---------------
4 files changed, 22 insertions(+), 19 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/hisilicon-hns-dsaf.txt b/Documentation/devicetree/bindings/net/hisilicon-hns-dsaf.txt
index c1358ea..72efe71 100644
--- a/Documentation/devicetree/bindings/net/hisilicon-hns-dsaf.txt
+++ b/Documentation/devicetree/bindings/net/hisilicon-hns-dsaf.txt
@@ -42,7 +42,9 @@ Required properties:
- cpld-syscon: is syscon handle for cpld register. It is not required if there
isn't cpld device.
- cpld-ctrl-reg: is cpld register offset. It is not required if there isn't
- cpld-syscon.
+ cpld-syscon.
+- port-rst-offset: is offset of reset field for each port in dsaf. Its value
+ depends on the hardware user manual.
[1] Documentation/devicetree/bindings/net/phy.txt
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c
index f1a477d..190477e 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c
+++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c
@@ -664,6 +664,7 @@ static int hns_mac_get_info(struct hns_mac_cb *mac_cb)
mac_cb->max_frm = MAC_DEFAULT_MTU;
mac_cb->tx_pause_frm_time = MAC_DEFAULT_PAUSE_TIME;
+ mac_cb->port_rst_off = mac_cb->mac_id;
/* if the dsaf node doesn't contain a port subnode,
* parse the old dts to get phy-handle from dsaf node
@@ -694,6 +695,15 @@ static int hns_mac_get_info(struct hns_mac_cb *mac_cb)
}
mac_cb->serdes_ctrl = syscon;
+ ret = fwnode_property_read_u32(mac_cb->fw_port,
+ "port-rst-offset",
+ &mac_cb->port_rst_off);
+ if (ret) {
+ dev_dbg(mac_cb->dev,
+ "mac%d port-rst-offset not found, use default value.\n",
+ mac_cb->mac_id);
+ }
+
syscon = syscon_node_to_regmap(
of_parse_phandle(to_of_node(mac_cb->fw_port),
"cpld-syscon", 0));
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h
index 719816b..7be7104 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h
+++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h
@@ -318,6 +318,7 @@ struct hns_mac_cb {
struct regmap *serdes_ctrl;
struct regmap *cpld_ctrl;
u32 cpld_ctrl_reg;
+ u32 port_rst_off;
struct mac_entry_idx addr_entry_idx[DSAF_MAX_VM_NUM];
u8 sfp_prsnt;
u8 cpld_led_value;
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c
index 7e52ab5..fc3056e 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c
+++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c
@@ -136,11 +136,7 @@ void hns_dsaf_xge_srst_by_port(struct dsaf_device *dsaf_dev, u32 port, u32 val)
return;
reg_val |= RESET_REQ_OR_DREQ;
-
- if (!HNS_DSAF_IS_DEBUG(dsaf_dev))
- reg_val |= 0x2082082 << port;
- else
- reg_val |= 0x2082082 << (dsaf_dev->reset_offset + 6);
+ reg_val |= 0x2082082 << dsaf_dev->mac_cb[port]->port_rst_off;
if (val == 0)
reg_addr = DSAF_SUB_SC_XGE_RESET_REQ_REG;
@@ -159,11 +155,7 @@ void hns_dsaf_xge_core_srst_by_port(struct dsaf_device *dsaf_dev,
if (port >= DSAF_XGE_NUM)
return;
- if (!HNS_DSAF_IS_DEBUG(dsaf_dev))
- reg_val |= XGMAC_TRX_CORE_SRST_M << port;
- else
- reg_val |= XGMAC_TRX_CORE_SRST_M <<
- (dsaf_dev->reset_offset + 6);
+ reg_val |= XGMAC_TRX_CORE_SRST_M << dsaf_dev->mac_cb[port]->port_rst_off;
if (val == 0)
reg_addr = DSAF_SUB_SC_XGE_RESET_REQ_REG;
@@ -177,17 +169,19 @@ void hns_dsaf_ge_srst_by_port(struct dsaf_device *dsaf_dev, u32 port, u32 val)
{
u32 reg_val_1;
u32 reg_val_2;
+ u32 port_rst_off;
if (port >= DSAF_GE_NUM)
return;
if (!HNS_DSAF_IS_DEBUG(dsaf_dev)) {
reg_val_1 = 0x1 << port;
+ port_rst_off = dsaf_dev->mac_cb[port]->port_rst_off;
/* there is difference between V1 and V2 in register.*/
if (AE_IS_VER1(dsaf_dev->dsaf_ver))
- reg_val_2 = 0x1041041 << port;
+ reg_val_2 = 0x1041041 << port_rst_off;
else
- reg_val_2 = 0x2082082 << port;
+ reg_val_2 = 0x2082082 << port_rst_off;
if (val == 0) {
dsaf_write_sub(dsaf_dev, DSAF_SUB_SC_GE_RESET_REQ1_REG,
@@ -226,12 +220,8 @@ void hns_ppe_srst_by_port(struct dsaf_device *dsaf_dev, u32 port, u32 val)
{
u32 reg_val = 0;
u32 reg_addr;
-
- if (!HNS_DSAF_IS_DEBUG(dsaf_dev))
- reg_val |= RESET_REQ_OR_DREQ << port;
- else
- reg_val |= RESET_REQ_OR_DREQ <<
- (dsaf_dev->reset_offset + 6);
+
+ reg_val |= RESET_REQ_OR_DREQ << dsaf_dev->mac_cb[port]->port_rst_off;
if (val == 0)
reg_addr = DSAF_SUB_SC_PPE_RESET_REQ_REG;
--
1.9.1
^ permalink raw reply related
* Re: [PATCH net-next 00/10] net: hns: add support of debug dsaf device
From: Yisen Zhuang @ 2016-04-28 6:47 UTC (permalink / raw)
To: devicetree-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
Cc: mark.rutland-5wv7dgnIgG8, pawel.moll-5wv7dgnIgG8,
ijc+devicetree-KcIKpvwj1kUDXYZnReoRVg,
catalin.marinas-5wv7dgnIgG8, will.deacon-5wv7dgnIgG8,
linuxarm-hv44wF8Li93QT0dZR+AlfA, robh+dt-DgEjT+Ai2ygdnm+yROfE0A,
galak-sgV2jX0FEOL9JmXXK+q4OQ, davem-fT/PcQaiUtIeIZ0/mPfg9Q
In-Reply-To: <1461826902-51368-1-git-send-email-Yisen.Zhuang-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
I am sorry for my misoperation, please ignore this series, thanks.
在 2016/4/28 15:01, Yisen Zhuang 写道:
> There are two kinds of dsaf device in hns, one is for service ports,
> contains crossbar in it, can work under different mode. Another is for
> debug port, only can work under single port mode. The current code only
> declared a dsaf device for both service ports and debug ports.It is not so
> readability. This patch separate it to three platform devices to make the
> code simple and more readability.
>
> all port in one platform device(old):
> CPU
> |
> | DSAF(one platform device)
> -------------------------------------------------------------- /
> | | | | | /
> | PPE PPE PPE | /
> | | | | | /
> | | | | | /
> | crossbar | | | /
> | | | | |/
> | ----------------------------------- | | |
> | | | | | | | | | |
> | | | | | | | | | |
> | MAC MAC MAC MAC MAC MAC MAC MAC |
> | | | | | | | | | |
> --------------------------------------------------------------
> | | | | | | | |
> PHY PHY PHY PHY PHY PHY PHY PHY
>
> separate all ports to three platform(new):
> CPU
> |
> -----------------------------------
> | | |
> ---------------------------------------------- --------- ---------
> | | | | | | | |
> | PPE | | PPE | | PPE |
> | | | | | | | | |
> | | | | | | | | |
> | crossbar | | | | | | |
> | | | | | | | | |
> | ---------------------------------- | | | | | | |
> | | | | | | | | | | | | | |
> | | | | | | | | | | | | | |
> | MAC MAC MAC MAC MAC MAC | | MAC | | MAC |
> | | | | | | | | | | | | | |
> ---------------------------------------------- --------- ---------
> | | | | | | \ / | / |
> PHY PHY PHY PHY PHY PHY \ / PHY / PHY
> \ / /
> \ / /
> DSAF(three platform device)
>
> I have tested this patchset on d02 board, it is compatible with the dts in
> d02 board.
>
> For more details, please see individual patches.
>
> MBR.
>
> Daode Huang (3):
> net: hns: set debug port irq index to 0
> net: hns: add syscon operation for dsaf
> net: hns: sort the header file by alphabetical order
>
> Yisen Zhuang (6):
> net: hns: add attribute port-idx-in-ae in enet node.
> net: hns: add attribute reset-field-offset for dsaf node
> net: hns: separate debug dsaf device from service dsaf device
> net: hns: add attribute cpld_ctrl for dsaf port node
> net: hns: add attribute port-rst-offset for dsaf port node
> net: hns: add attribute port-mode-offset for dsaf port node
>
> huangdaode (1):
> net: hns: add a new dsaf mode for debug port
>
> .../devicetree/bindings/net/hisilicon-hns-dsaf.txt | 60 +++++-
> .../devicetree/bindings/net/hisilicon-hns-nic.txt | 30 ++-
> drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c | 73 ++-----
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c | 231 ++++++++++++++-------
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.h | 12 +-
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_main.c | 141 ++++++++-----
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_main.h | 16 +-
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_misc.c | 173 +++++++++------
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_ppe.c | 51 +----
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_ppe.h | 1 -
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_rcb.c | 187 +++++++----------
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_rcb.h | 2 +-
> drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h | 38 ++--
> drivers/net/ethernet/hisilicon/hns/hns_enet.c | 17 +-
> drivers/net/ethernet/hisilicon/hns/hns_enet.h | 3 +
> 15 files changed, 584 insertions(+), 451 deletions(-)
>
--
To unsubscribe from this list: send the line "unsubscribe devicetree" 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 3/3] gre: receive also TEB packets for lwtunnels
From: Simon Horman @ 2016-04-28 6:49 UTC (permalink / raw)
To: Jiri Benc; +Cc: netdev, Pravin B Shelar, Thomas Graf
In-Reply-To: <017d8f2b1c00e8fdbdb92a5898b1b5f365b58d6d.1461346798.git.jbenc@redhat.com>
On Fri, Apr 22, 2016 at 07:44:08PM +0200, Jiri Benc wrote:
> For ipgre interfaces in collect metadata mode, receive also traffic with
> encapsulated Ethernet headers. The lwtunnel users are supposed to sort this
> out correctly. This allows to have mixed Ethernet + L3-only traffic on the
> same lwtunnel interface.
>
> To keep backwards compatibility and prevent any surprises, gretap interfaces
> have priority in receiving packets with Ethernet headers.
Hi Jiri,
I have had some success wiring up Open vSwitch to use this patch for
transmit. However, I am wondering if something more is needed to allow
differentiation between packets with and without an L2 header present
on receive.
I had luck getting receive working with the following:
From: Simon Horman <simon.horman@netronome.com>
Date: Mon, 18 Apr 2016 17:48:47 +1000
Subject: [PATCH] gre: mark presense of l2 when recieving TEB packets for lwtunnels
There seems to be some way for receivers to differentiate between
packets recieved with and without an l2 header. The approach taken here
is to use a new mode bit in struct ip_tunnel_key.
Another approach might be to store tpi->proto in tunnel metadata,
though that would consume 16 bits somewhere and seems like overkill
at this point.
Signed-off-by: Simon Horman <simon.horman@netronome.com>
---
include/net/ip_tunnels.h | 6 ++++++
net/ipv4/ip_gre.c | 3 +++
2 files changed, 9 insertions(+)
diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h
index d916b4315903..cdf71ced429e 100644
--- a/include/net/ip_tunnels.h
+++ b/include/net/ip_tunnels.h
@@ -58,6 +58,12 @@ struct ip_tunnel_key {
/* Flags for ip_tunnel_info mode. */
#define IP_TUNNEL_INFO_TX 0x01 /* represents tx tunnel parameters */
#define IP_TUNNEL_INFO_IPV6 0x02 /* key contains IPv6 addresses */
+#define IP_TUNNEL_INFO_L2_PRESENT 0x04 /* Set on receive by tunnels that
+ * may receive packets both with
+ * and without an L2 header present
+ * when a packet is received with
+ * L2 header present.
+ */
/* Maximum tunnel options length. */
#define IP_TUNNEL_OPTS_MAX \
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 0a4af2896a15..1290695fbc95 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -401,6 +401,9 @@ static int __ipgre_rcv(struct sk_buff *skb, const struct tnl_ptk_info *tpi,
tun_dst = ip_tun_rx_dst(skb, flags, tun_id, 0);
if (!tun_dst)
return PACKET_REJECT;
+ if (tpi->proto == htons(ETH_P_TEB))
+ tun_dst->u.tun_info.mode |=
+ IP_TUNNEL_INFO_L2_PRESENT;
}
ip_tunnel_rcv(tunnel, skb, tpi, tun_dst, log_ecn_error);
--
2.7.0.rc3.207.g0ac5344
^ permalink raw reply related
* Re: [PATCH v2 0/2] pegasus: correct buffer sizes
From: Johannes Berg @ 2016-04-28 6:49 UTC (permalink / raw)
To: David Laight, Petko Manolov
Cc: netdev@vger.kernel.org, davem@davemloft.net, a1291762@gmail.com
In-Reply-To: <063D6719AE5E284EB5DD2968C1650D6D5F4ABB06@AcuExch.aculab.com>
On Wed, 2016-04-27 at 16:26 +0000, David Laight wrote:
> From: Johannes Berg
> >
> > Sent: 27 April 2016 10:44
> > On Wed, 2016-04-27 at 12:33 +0300, Petko Manolov wrote:
> > >
> > >
> > > Your guess turned out to be not so wild.;)All Pegasus devices are
> > > configured (by the driver) to append CRC at the end of each RX
> > > packet.However, the driver reports packet length that does not
> > > include it.
> > Interesting, then my guess was wrong though, since the length is
> > reported without it, or am I misunderstanding this?
> ...
>
> It is even possible that the crc is written into the rx buffer even
> though the length the hardware reports excludes it.
>
Right. I think a proper test would be to construct some kind of
ethernet-only frames, and check that they come out properly on the
other side. IP always has its own length field, so may be happy with
trailing garbage reported to the network stack, but doing raw ethernet
will let you check that the frame length is correct.
johannes
^ permalink raw reply
* [PATCH net-next 03/10] net: hns: add attribute port-idx-in-ae in enet node.
From: Yisen Zhuang @ 2016-04-28 7:01 UTC (permalink / raw)
To: devicetree, netdev, linux-arm-kernel
Cc: robh+dt, pawel.moll, mark.rutland, ijc+devicetree, galak, davem,
will.deacon, catalin.marinas, yankejian, huangdaode, salil.mehta,
lipeng321, liguozhu, xieqianqian, xuwei5, linuxarm
In-Reply-To: <1461826902-51368-1-git-send-email-Yisen.Zhuang@huawei.com>
This patch parse port-idx-in-ae in enet node.
port-idx-in-ae: is the index of port provided by AE.
In NIC mode of DSAF, all 6 PHYs of service DSAF are taken as ethernet
ports to the CPU. The port-idx-in-ae can be 0 to 5. Here is the diagram:
+-----+---------------+
| CPU |
+-+-+-+---+-+-+-+-+-+-+
| | | | | | | |
debug debug service
port port port
(0) (0) (0-5)
In Switch mode of DSAF, all 6 PHYs of service DSAF are taken as physical
ports connect to a LAN Switch while the CPU side assume itself have one
single NIC connect to this switch. In this case, the port-idx-in-ae will
be 0 only.
+-----+-----+------+------+
| CPU |
+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | service| port(0)
debug debug +------------+
port port | switch |
(0) (0) +-+-+-+-+-+-++
| | | | | |
external port
when port-idx-in-ae is not exists, old attribute port-id will be used
(only for compatible purpose, not recommended to use port-id in new code).
Signed-off-by: Daode Huang <huangdaode@hisilicon.com>
Signed-off-by: Yuzeng Zhuang <yisen.zhuang@huawei.com>
---
.../devicetree/bindings/net/hisilicon-hns-nic.txt | 30 +++++++++++++++++++-
drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c | 33 ++++------------------
drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h | 1 -
drivers/net/ethernet/hisilicon/hns/hns_enet.c | 17 ++++++++---
drivers/net/ethernet/hisilicon/hns/hns_enet.h | 3 ++
5 files changed, 51 insertions(+), 33 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/hisilicon-hns-nic.txt b/Documentation/devicetree/bindings/net/hisilicon-hns-nic.txt
index e6a9d1c..965f6be 100644
--- a/Documentation/devicetree/bindings/net/hisilicon-hns-nic.txt
+++ b/Documentation/devicetree/bindings/net/hisilicon-hns-nic.txt
@@ -36,6 +36,34 @@ Required properties:
| | | | | |
external port
+ This attribute is remained for compatible purpose. It is not recommended to
+ use it in new code.
+
+- port-idx-in-ae: is the index of port provided by AE.
+ In NIC mode of DSAF, all 6 PHYs of service DSAF are taken as ethernet ports
+ to the CPU. The port-idx-in-ae can be 0 to 5. Here is the diagram:
+ +-----+---------------+
+ | CPU |
+ +-+-+-+---+-+-+-+-+-+-+
+ | | | | | | | |
+ debug debug service
+ port port port
+ (0) (0) (0-5)
+
+ In Switch mode of DSAF, all 6 PHYs of service DSAF are taken as physical
+ ports connect to a LAN Switch while the CPU side assume itself have one
+ single NIC connect to this switch. In this case, the port-idx-in-ae will
+ be 0 only.
+ +-----+-----+------+------+
+ | CPU |
+ +-+-+-+-+-+-+-+-+-+-+-+-+-+
+ | | service| port(0)
+ debug debug +------------+
+ port port | switch |
+ (0) (0) +-+-+-+-+-+-++
+ | | | | | |
+ external port
+
- local-mac-address: mac addr of the ethernet interface
Example:
@@ -43,6 +71,6 @@ Example:
ethernet@0{
compatible = "hisilicon,hns-nic-v1";
ae-handle = <&dsaf0>;
- port-id = <0>;
+ port-idx-in-ae = <0>;
local-mac-address = [a2 14 e4 4b 56 76];
};
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c b/drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c
index 1e8bf22..1c86336 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c
+++ b/drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c
@@ -29,25 +29,6 @@ static struct hns_mac_cb *hns_get_mac_cb(struct hnae_handle *handle)
return vf_cb->mac_cb;
}
-/**
- * hns_ae_map_eport_to_dport - translate enet port id to dsaf port id
- * @port_id: enet port id
- *: debug port 0-1, service port 2 -7 (dsaf mode only 2)
- * return: dsaf port id
- *: service ports 0 - 5, debug port 6-7
- **/
-static int hns_ae_map_eport_to_dport(u32 port_id)
-{
- int port_index;
-
- if (port_id < DSAF_DEBUG_NW_NUM)
- port_index = port_id + DSAF_SERVICE_PORT_NUM_PER_DSAF;
- else
- port_index = port_id - DSAF_DEBUG_NW_NUM;
-
- return port_index;
-}
-
static struct dsaf_device *hns_ae_get_dsaf_dev(struct hnae_ae_dev *dev)
{
return container_of(dev, struct dsaf_device, ae_dev);
@@ -110,7 +91,6 @@ static struct ring_pair_cb *hns_ae_get_ring_pair(struct hnae_queue *q)
struct hnae_handle *hns_ae_get_handle(struct hnae_ae_dev *dev,
u32 port_id)
{
- int port_idx;
int vfnum_per_port;
int qnum_per_vf;
int i;
@@ -120,11 +100,10 @@ struct hnae_handle *hns_ae_get_handle(struct hnae_ae_dev *dev,
struct hnae_vf_cb *vf_cb;
dsaf_dev = hns_ae_get_dsaf_dev(dev);
- port_idx = hns_ae_map_eport_to_dport(port_id);
- ring_pair_cb = hns_ae_get_base_ring_pair(dsaf_dev, port_idx);
- vfnum_per_port = hns_ae_get_vf_num_per_port(dsaf_dev, port_idx);
- qnum_per_vf = hns_ae_get_q_num_per_vf(dsaf_dev, port_idx);
+ ring_pair_cb = hns_ae_get_base_ring_pair(dsaf_dev, port_id);
+ vfnum_per_port = hns_ae_get_vf_num_per_port(dsaf_dev, port_id);
+ qnum_per_vf = hns_ae_get_q_num_per_vf(dsaf_dev, port_id);
vf_cb = kzalloc(sizeof(*vf_cb) +
qnum_per_vf * sizeof(struct hnae_queue *), GFP_KERNEL);
@@ -163,14 +142,14 @@ struct hnae_handle *hns_ae_get_handle(struct hnae_ae_dev *dev,
}
vf_cb->dsaf_dev = dsaf_dev;
- vf_cb->port_index = port_idx;
- vf_cb->mac_cb = &dsaf_dev->mac_cb[port_idx];
+ vf_cb->port_index = port_id;
+ vf_cb->mac_cb = &dsaf_dev->mac_cb[port_id];
ae_handle->phy_if = vf_cb->mac_cb->phy_if;
ae_handle->phy_node = vf_cb->mac_cb->phy_node;
ae_handle->if_support = vf_cb->mac_cb->if_support;
ae_handle->port_type = vf_cb->mac_cb->mac_type;
- ae_handle->dport_id = port_idx;
+ ae_handle->dport_id = port_id;
return ae_handle;
vf_id_err:
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h
index 8782608..ed0043a 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h
+++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_reg.h
@@ -24,7 +24,6 @@
#define DSAF_SERVICE_NW_NUM 6
#define DSAF_COMM_CHN DSAF_SERVICE_NW_NUM
#define DSAF_GE_NUM ((DSAF_SERVICE_NW_NUM) + (DSAF_DEBUG_NW_NUM))
-#define DSAF_PORT_NUM ((DSAF_SERVICE_NW_NUM) + (DSAF_DEBUG_NW_NUM))
#define DSAF_XGE_NUM DSAF_SERVICE_NW_NUM
#define DSAF_PORT_TYPE_NUM 3
#define DSAF_NODE_NUM 18
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_enet.c b/drivers/net/ethernet/hisilicon/hns/hns_enet.c
index 687204b..e47aff2 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns/hns_enet.c
@@ -1873,6 +1873,7 @@ static int hns_nic_dev_probe(struct platform_device *pdev)
struct net_device *ndev;
struct hns_nic_priv *priv;
struct device_node *node = dev->of_node;
+ u32 port_id;
int ret;
ndev = alloc_etherdev_mq(sizeof(struct hns_nic_priv), NIC_MAX_Q_PER_VF);
@@ -1896,10 +1897,18 @@ static int hns_nic_dev_probe(struct platform_device *pdev)
dev_err(dev, "not find ae-handle\n");
goto out_read_prop_fail;
}
-
- ret = of_property_read_u32(node, "port-id", &priv->port_id);
- if (ret)
- goto out_read_prop_fail;
+ /* try to find port-idx-in-ae first */
+ ret = of_property_read_u32(node, "port-idx-in-ae", &port_id);
+ if (ret) {
+ /* only for old code compatible */
+ ret = of_property_read_u32(node, "port-id", &port_id);
+ if (ret)
+ goto out_read_prop_fail;
+ /* for old dts, we need to caculate the port offset */
+ port_id = port_id < HNS_SRV_OFFSET ? port_id + HNS_DEBUG_OFFSET
+ : port_id - HNS_SRV_OFFSET;
+ }
+ priv->port_id = port_id;
hns_init_mac_addr(ndev);
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_enet.h b/drivers/net/ethernet/hisilicon/hns/hns_enet.h
index c68ab3d..337efa5 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_enet.h
+++ b/drivers/net/ethernet/hisilicon/hns/hns_enet.h
@@ -18,6 +18,9 @@
#include "hnae.h"
+#define HNS_DEBUG_OFFSET 6
+#define HNS_SRV_OFFSET 2
+
enum hns_nic_state {
NIC_STATE_TESTING = 0,
NIC_STATE_RESETTING,
--
1.9.1
^ permalink raw reply related
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