* [PATCH net-next 3/7] be2net: don't limit max MAC and VLAN counts
From: Sathya Perla @ 2013-08-06 3:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <1375761441-22343-1-git-send-email-sathya.perla@emulex.com>
From: Vasundhara Volam <vasundhara.volam@emulex.com>
For SH-R and Lancer-R, use the FW supported values for Max unicast MACs,
Max VLANs and Max multicast MACs.
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com>
Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 7 -------
1 files changed, 0 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 834f77b..dc45e1a 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2995,13 +2995,6 @@ static void be_get_resources(struct be_adapter *adapter)
}
if (profile_present) {
- /* Sanity fixes for Lancer */
- adapter->max_pmac_cnt = min_t(u16, adapter->max_pmac_cnt,
- BE_UC_PMAC_COUNT);
- adapter->max_vlans = min_t(u16, adapter->max_vlans,
- BE_NUM_VLANS_SUPPORTED);
- adapter->max_mcast_mac = min_t(u16, adapter->max_mcast_mac,
- BE_MAX_MC);
adapter->max_tx_queues = min_t(u16, adapter->max_tx_queues,
MAX_TX_QS);
adapter->max_rss_queues = min_t(u16, adapter->max_rss_queues,
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 2/7] be2net: Do not call get_die_temperature cmd for VF
From: Sathya Perla @ 2013-08-06 3:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <1375761441-22343-1-git-send-email-sathya.perla@emulex.com>
From: Vasundhara Volam <vasundhara.volam@emulex.com>
This is a chip wide value and the PFs already report it.
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com>
Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 3df1503..834f77b 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -4143,7 +4143,8 @@ static void be_worker(struct work_struct *work)
be_cmd_get_stats(adapter, &adapter->stats_cmd);
}
- if (MODULO(adapter->work_counter, adapter->be_get_temp_freq) == 0)
+ if (be_physfn(adapter) &&
+ MODULO(adapter->work_counter, adapter->be_get_temp_freq) == 0)
be_cmd_get_die_temperature(adapter);
for_all_rx_queues(adapter, rxo, i) {
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 1/7] be2net: Adding more speeds reported by get_settings
From: Sathya Perla @ 2013-08-06 3:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <1375761441-22343-1-git-send-email-sathya.perla@emulex.com>
From: Vasundhara Volam <vasundhara.volam@emulex.com>
The new speeds are supported by variants of the Skyhawk-R chip.
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com>
Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
---
drivers/net/ethernet/emulex/benet/be_cmds.c | 6 ++++++
drivers/net/ethernet/emulex/benet/be_cmds.h | 5 ++++-
2 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 613d887..bef25b7 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -1464,6 +1464,12 @@ static int be_mac_to_link_speed(int mac_speed)
return 1000;
case PHY_LINK_SPEED_10GBPS:
return 10000;
+ case PHY_LINK_SPEED_20GBPS:
+ return 20000;
+ case PHY_LINK_SPEED_25GBPS:
+ return 25000;
+ case PHY_LINK_SPEED_40GBPS:
+ return 40000;
}
return 0;
}
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index eb541f0..9b91608 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -960,7 +960,10 @@ enum {
PHY_LINK_SPEED_10MBPS = 0x1,
PHY_LINK_SPEED_100MBPS = 0x2,
PHY_LINK_SPEED_1GBPS = 0x3,
- PHY_LINK_SPEED_10GBPS = 0x4
+ PHY_LINK_SPEED_10GBPS = 0x4,
+ PHY_LINK_SPEED_20GBPS = 0x5,
+ PHY_LINK_SPEED_25GBPS = 0x6,
+ PHY_LINK_SPEED_40GBPS = 0x7
};
struct be_cmd_resp_link_status {
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 0/7] be2net patch set
From: Sathya Perla @ 2013-08-06 3:57 UTC (permalink / raw)
To: netdev
The following fix/cleanup patches are for the net-next tree. Pls apply.
Thanks!
Sathya Perla (1):
be2net: update driver version
Vasundhara Volam (6):
be2net: Adding more speeds reported by get_settings
be2net: Do not call get_die_temperature cmd for VF
be2net: don't limit max MAC and VLAN counts
be2net: Fix displaying supported speeds for BE2
be2net: fixup log msgs for async events
be2net: Initialize "status" in be_cmd_get_die_temperature()
drivers/net/ethernet/emulex/benet/be.h | 2 +-
drivers/net/ethernet/emulex/benet/be_cmds.c | 20 +++++++++++++++++---
drivers/net/ethernet/emulex/benet/be_cmds.h | 5 ++++-
drivers/net/ethernet/emulex/benet/be_main.c | 10 ++--------
4 files changed, 24 insertions(+), 13 deletions(-)
^ permalink raw reply
* macvtap MAX_SKB_FRAGS fix...
From: David Miller @ 2013-08-06 3:51 UTC (permalink / raw)
To: netdev; +Cc: jasowang, mst
Jason, I want to get that into v3.4.x -stable but it needs a backport
and I'd rather someone like you who knows the code does it rather
than I.
The issue seems that be that the arguments to uarg->callback() were
different back then.
Thanks.
^ permalink raw reply
* [PATCH] tcp: cubic: fix bug in bictcp_acked()
From: Eric Dumazet @ 2013-08-06 3:05 UTC (permalink / raw)
To: David Miller; +Cc: Neal Cardwell, netdev, Van Jacobson, Yuchung Cheng
From: Eric Dumazet <edumazet@google.com>
While investigating about strange increase of retransmit rates
on hosts ~24 days after boot, Van found hystart was disabled
if ca->epoch_start was 0, as following condition is true
when tcp_time_stamp high order bit is set.
(s32)(tcp_time_stamp - ca->epoch_start) < HZ
Quoting Van :
At initialization & after every loss ca->epoch_start is set to zero so
I believe that the above line will turn off hystart as soon as the 2^31
bit is set in tcp_time_stamp & hystart will stay off for 24 days.
I think we've observed that cubic's restart is too aggressive without
hystart so this might account for the higher drop rate we observe.
Diagnosed-by: Van Jacobson <vanj@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
---
net/ipv4/tcp_cubic.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index a9077f4..2b8671d 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -414,7 +414,7 @@ static void bictcp_acked(struct sock *sk, u32 cnt, s32 rtt_us)
return;
/* Discard delay samples right after fast recovery */
- if ((s32)(tcp_time_stamp - ca->epoch_start) < HZ)
+ if (ca->epoch_start && (s32)(tcp_time_stamp - ca->epoch_start) < HZ)
return;
delay = (rtt_us << 3) / USEC_PER_MSEC;
^ permalink raw reply related
* Re: [PATCH v2.35 5/6] lib: Push MPLS tags in the OpenFlow 1.3 ordering
From: Joe Stringer @ 2013-08-06 2:04 UTC (permalink / raw)
To: Ben Pfaff
Cc: dev-yBygre7rU0TnMu66kgdUjQ@public.gmane.org, Ravi K,
netdev-u79uwXL29TY76Z2rM5mHXA, Isaku Yamahata
In-Reply-To: <1374289623-17056-6-git-send-email-horms-/R6kz+dDXgpPR4JQBCEnsQ@public.gmane.org>
[-- Attachment #1.1: Type: text/plain, Size: 23489 bytes --]
Hi Ben,
I realise that you've agreed to let Jesse review the datapath patch first,
but I wondered if I could get a little broad feedback on the approach on
the userspace side.
The general background to this patch is that we aim to keep the datapath
interface for MPLS actions simple. As such, when pushing an MPLS tag, this
is done immediately after the ethernet header regardless of the presence of
VLANs (As per OF1.3 spec). We can then implement OF1.2 behaviour by popping
any existing VLAN tags before applying MPLS actions, then pushing the tags
back on afterwards. This is all done on the odp actions translation side.
The patches leading up to this one allow us to store which protocol
inserted the MPLS action in the ofpact->compat field, and we translate VLAN
actions differently based on this value. There may be alternative
approaches, for instance having a per-bridge "MPLS behaviour" configuration
flag, set when the OpenFlow connection is established.
I'm wondering if this general approach is the preferred method, and as an
extension, what are your thoughts on how these patches add the new
'vlan_tci' field into 'struct xlate_in' and apply VLAN actions twice
(before and/or after MPLS actions). This provides a mechanism to support
both OF1.2 and OF1.3 behaviours within the scope of current VLAN support,
but may or may not fit in well with future ideas for QinQ (ie,
double-stacked 0x8100 tags) or 802.1ad support.
Cheers,
Joe
On Sat, Jul 20, 2013 at 12:07 PM, Simon Horman <horms-/R6kz+dDXgpPR4JQBCEnsQ@public.gmane.org> wrote:
> From: Joe Stringer <joe-Q1GJJQv1iO6lP80pJB477g@public.gmane.org>
>
> This patch modifies the push_mpls behaviour to follow the OpenFlow 1.3
> specification in the presence of VLAN tagged packets. From the spec:
>
> "Newly pushed tags should always be inserted as the outermost tag in the
> outermost valid location for that tag. When a new VLAN tag is pushed, it
> should be the outermost tag inserted, immediately after the Ethernet
> header and before other tags. Likewise, when a new MPLS tag is pushed,
> it should be the outermost tag inserted, immediately after the Ethernet
> header and before other tags."
>
> When the push_mpls action was inserted using OpenFlow 1.2, we implement
> the previous behaviour by inserting VLAN actions around the MPLS action
> in the odp translation; Pop VLAN tags before committing MPLS actions,
> and push the expected VLAN tag afterwards. The trigger condition for
> this is based on the ofpact->compat field.
>
> Signed-off-by: Joe Stringer <joe-Q1GJJQv1iO6lP80pJB477g@public.gmane.org>
> Signed-off-by: Simon Horman <horms-/R6kz+dDXgpPR4JQBCEnsQ@public.gmane.org>
> ---
> lib/flow.c | 2 +-
> lib/packets.c | 10 +-
> lib/packets.h | 2 +-
> ofproto/ofproto-dpif-xlate.c | 10 +-
> tests/ofproto-dpif.at | 237
> ++++++++++++++++++++++++++++++++++++++++++
> 5 files changed, 253 insertions(+), 8 deletions(-)
>
> diff --git a/lib/flow.c b/lib/flow.c
> index d899d26..3dc4489 100644
> --- a/lib/flow.c
> +++ b/lib/flow.c
> @@ -1035,7 +1035,7 @@ flow_compose(struct ofpbuf *b, const struct flow
> *flow)
> }
>
> if (eth_type_mpls(flow->dl_type)) {
> - b->l2_5 = b->l3;
> + b->l2_5 = (char*)b->l2 + ETH_HEADER_LEN;
> push_mpls(b, flow->dl_type, flow->mpls_lse);
> }
> }
> diff --git a/lib/packets.c b/lib/packets.c
> index 7fe6513..3bc155f 100644
> --- a/lib/packets.c
> +++ b/lib/packets.c
> @@ -240,11 +240,11 @@ eth_mpls_depth(const struct ofpbuf *packet)
>
> /* Set ethertype of the packet. */
> void
> -set_ethertype(struct ofpbuf *packet, ovs_be16 eth_type)
> +set_ethertype(struct ofpbuf *packet, ovs_be16 eth_type, bool inner)
> {
> struct eth_header *eh = packet->data;
>
> - if (eh->eth_type == htons(ETH_TYPE_VLAN)) {
> + if (inner && eh->eth_type == htons(ETH_TYPE_VLAN)) {
> ovs_be16 *p;
> p = (ovs_be16 *)((char *)(packet->l2_5 ? packet->l2_5 :
> packet->l3) - 2);
> *p = eth_type;
> @@ -351,8 +351,8 @@ push_mpls(struct ofpbuf *packet, ovs_be16 ethtype,
> ovs_be32 lse)
>
> if (!is_mpls(packet)) {
> /* Set ethtype and MPLS label stack entry. */
> - set_ethertype(packet, ethtype);
> - packet->l2_5 = packet->l3;
> + set_ethertype(packet, ethtype, false);
> + packet->l2_5 = (char*)packet->l2 + ETH_HEADER_LEN;
> }
>
> /* Push new MPLS shim header onto packet. */
> @@ -373,7 +373,7 @@ pop_mpls(struct ofpbuf *packet, ovs_be16 ethtype)
> size_t len;
> mh = packet->l2_5;
> len = (char*)packet->l2_5 - (char*)packet->l2;
> - set_ethertype(packet, ethtype);
> + set_ethertype(packet, ethtype, true);
> if (mh->mpls_lse & htonl(MPLS_BOS_MASK)) {
> packet->l2_5 = NULL;
> } else {
> diff --git a/lib/packets.h b/lib/packets.h
> index e852761..923fd53 100644
> --- a/lib/packets.h
> +++ b/lib/packets.h
> @@ -142,7 +142,7 @@ void eth_pop_vlan(struct ofpbuf *);
>
> uint16_t eth_mpls_depth(const struct ofpbuf *packet);
>
> -void set_ethertype(struct ofpbuf *packet, ovs_be16 eth_type);
> +void set_ethertype(struct ofpbuf *packet, ovs_be16 eth_type, bool inner);
>
> const char *eth_from_hex(const char *hex, struct ofpbuf **packetp);
> void eth_format_masked(const uint8_t eth[ETH_ADDR_LEN],
> diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
> index 8a0c32e..457a1d1 100644
> --- a/ofproto/ofproto-dpif-xlate.c
> +++ b/ofproto/ofproto-dpif-xlate.c
> @@ -1912,6 +1912,12 @@ tunnel_ecn_ok(struct xlate_ctx *ctx)
> return true;
> }
>
> +static bool
> +mpls_compat_behaviour(enum ofputil_action_code compat)
> +{
> + return (compat != OFPUTIL_OFPAT13_PUSH_MPLS);
> +}
> +
> static void
> vlan_tci_restore(struct xlate_in *xin, ovs_be16 *tci_ptr, ovs_be16
> orig_tci)
> {
> @@ -2089,7 +2095,9 @@ do_xlate_actions(const struct ofpact *ofpacts,
> size_t ofpacts_len,
>
> /* Save and pop any existing VLAN tags if running in OF1.2
> mode. */
> ctx->xin->vlan_tci = *vlan_tci;
> - flow->vlan_tci = htons(0);
> + if (mpls_compat_behaviour(a->compat)) {
> + flow->vlan_tci = htons(0);
> + }
> vlan_tci = &ctx->xin->vlan_tci;
> break;
>
> diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
> index 62531da..5e85ab2 100644
> --- a/tests/ofproto-dpif.at
> +++ b/tests/ofproto-dpif.at
> @@ -1017,6 +1017,243 @@ NXST_FLOW reply:
> OVS_VSWITCHD_STOP
> AT_CLEANUP
>
> +AT_SETUP([ofproto-dpif - OF1.3+ VLAN+MPLS handling])
> +OVS_VSWITCHD_START([dnl
> + add-port br0 p1 -- set Interface p1 type=dummy
> +])
> +ON_EXIT([kill `cat ovs-ofctl.pid`])
> +
> +AT_CAPTURE_FILE([ofctl_monitor.log])
> +AT_DATA([flows.txt], [dnl
> +cookie=0xa dl_src=40:44:44:44:55:44
> actions=push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],controller
> +cookie=0xa dl_src=40:44:44:44:55:45
> actions=push_vlan:0x8100,mod_vlan_vid:99,push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],controller
> +cookie=0xa dl_src=40:44:44:44:55:46
> actions=push_vlan:0x8100,mod_vlan_vid:99,push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],controller
> +cookie=0xa dl_src=40:44:44:44:55:47
> actions=push_vlan:0x8100,load:99->OXM_OF_VLAN_VID[[]],push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],controller
> +cookie=0xa dl_src=40:44:44:44:55:48
> actions=push_vlan:0x8100,load:99->OXM_OF_VLAN_VID[[]],push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],controller
> +cookie=0xa dl_src=40:44:44:44:55:49
> actions=push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],push_vlan:0x8100,mod_vlan_vid:99,controller
> +cookie=0xa dl_src=40:44:44:44:55:50
> actions=push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],push_vlan:0x8100,mod_vlan_vid:99,controller
> +cookie=0xa dl_src=40:44:44:44:55:51
> actions=push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],push_vlan:0x8100,load:99->OXM_OF_VLAN_VID[[]],mod_vlan_pcp:1,controller
> +cookie=0xa dl_src=40:44:44:44:55:52
> actions=push_mpls:0x8847,load:10->OXM_OF_MPLS_LABEL[[]],load:3->OXM_OF_MPLS_TC[[]],push_vlan:0x8100,load:99->OXM_OF_VLAN_VID[[]],mod_vlan_pcp:1,controller
> +])
> +AT_CHECK([ovs-ofctl --protocols=OpenFlow13 add-flows br0 flows.txt])
> +
> +dnl Modified MPLS controller action.
> +dnl The input packet has a VLAN tag, but because we push an MPLS tag in
> +dnl OF1.3 mode, we can no longer see it.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:44,dst=50:54:00:00:00:07),eth_type(0x8100),vlan(vid=88,pcp=7),encap(eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:44,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:44,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:44,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, we push a VLAN tag, then an MPLS tag in OF1.3 mode, so
> we
> +dnl can only see the MPLS tag in the result.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:45,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:45,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:45,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:45,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, the input packet is vlan-tagged; we update this tag then
> +dnl push an MPLS tag in OF1.3 mode. As such, we can only see the MPLS tag
> in
> +dnl the result.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:46,dst=50:54:00:00:00:07),eth_type(0x8100),vlan(vid=88,pcp=7),encap(eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:46,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:46,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:46,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, we push a VLAN tag, then an MPLS tag in OF1.3 mode, so
> we
> +dnl can only see the MPLS tag in the result.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:47,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:47,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:47,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:47,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, the input packet is vlan-tagged; we update this tag then
> +dnl push an MPLS tag in OF1.3 mode. As such, we can only see the MPLS tag
> in
> +dnl the result.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:48,dst=50:54:00:00:00:07),eth_type(0x8100),vlan(vid=88,pcp=7),encap(eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:48,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:48,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=64 in_port=1 (via action)
> data_len=64 (unbuffered)
>
> +mpls,metadata=0,in_port=0,vlan_tci=0x0000,dl_src=40:44:44:44:55:48,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, we push the MPLS tag before pushing a VLAN tag, so we
> see
> +dnl both of these in the final flow.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:49,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:49,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:49,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:49,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, the input packet in vlan-tagged, which should be
> stripped
> +dnl before we push the MPLS and VLAN tags.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:50,dst=50:54:00:00:00:07),eth_type(0x8100),vlan(vid=88,pcp=7),encap(eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:50,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:50,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=0,dl_src=40:44:44:44:55:50,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, we push the MPLS tag before pushing a VLAN tag, so we
> see
> +dnl both of these in the final flow.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:51,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:51,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:51,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:51,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +dnl Modified MPLS controller action.
> +dnl In this test, the input packet in vlan-tagged, which should be
> stripped
> +dnl before we push the MPLS and VLAN tags.
> +AT_CHECK([ovs-ofctl monitor br0 65534 -P nxm --detach --pidfile 2>
> ofctl_monitor.log])
> +
> +for i in 1 2 3; do
> + ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=40:44:44:44:55:52,dst=50:54:00:00:00:07),eth_type(0x8100),vlan(vid=88,pcp=7),encap(eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
> +done
> +OVS_WAIT_UNTIL([test `wc -l < ofctl_monitor.log` -ge 6])
> +ovs-appctl -t ovs-ofctl exit
> +
> +AT_CHECK([cat ofctl_monitor.log], [0], [dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:52,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:52,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +dnl
> +NXT_PACKET_IN (xid=0x0): cookie=0xa total_len=68 in_port=1 (via action)
> data_len=68 (unbuffered)
>
> +mpls,metadata=0,in_port=0,dl_vlan=99,dl_vlan_pcp=1,dl_src=40:44:44:44:55:52,dl_dst=50:54:00:00:00:07,mpls_label=10,mpls_tc=3,mpls_ttl=64,mpls_bos=1
> +])
> +
> +AT_CHECK([ovs-appctl time/warp 5000], [0], [ignore])
> +AT_CHECK([ovs-ofctl dump-flows br0 | ofctl_strip | sort], [0], [dnl
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:44
> actions=push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:45
> actions=mod_vlan_vid:99,push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:46
> actions=mod_vlan_vid:99,push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:47
> actions=load:0x63->OXM_OF_VLAN_VID[[]],push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:48
> actions=load:0x63->OXM_OF_VLAN_VID[[]],push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:49
> actions=push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],mod_vlan_vid:99,CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:50
> actions=push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],mod_vlan_vid:99,CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:51
> actions=push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],load:0x63->OXM_OF_VLAN_VID[[]],mod_vlan_pcp:1,CONTROLLER:65535
> + cookie=0xa, n_packets=3, n_bytes=180, dl_src=40:44:44:44:55:52
> actions=push_mpls:0x8847,load:0xa->OXM_OF_MPLS_LABEL[[]],load:0x3->OXM_OF_MPLS_TC[[]],load:0x63->OXM_OF_VLAN_VID[[]],mod_vlan_pcp:1,CONTROLLER:65535
> +NXST_FLOW reply:
> +])
> +
> +OVS_VSWITCHD_STOP
> +AT_CLEANUP
> +
> AT_SETUP([ofproto-dpif - fragment handling])
> OVS_VSWITCHD_START
> ADD_OF_PORTS([br0], [1], [2], [3], [4], [5], [6], [90])
> --
> 1.7.10.4
>
>
[-- Attachment #1.2: Type: text/html, Size: 25879 bytes --]
[-- Attachment #2: Type: text/plain, Size: 0 bytes --]
^ permalink raw reply
* RE: [net-next 07/15] ixgbe: fix SFF data dumps of SFP+ modules
From: Tantilov, Emil S @ 2013-08-06 0:56 UTC (permalink / raw)
To: Ben Hutchings, Kirsher, Jeffrey T
Cc: davem@davemloft.net, netdev@vger.kernel.org, gospo@redhat.com,
sassmann@redhat.com
In-Reply-To: <1375434517.24371.24.camel@deadeye.wl.decadent.org.uk>
>-----Original Message-----
>From: netdev-owner@vger.kernel.org [mailto:netdev-
>owner@vger.kernel.org] On Behalf Of Ben Hutchings
>Sent: Friday, August 02, 2013 2:09 AM
>To: Kirsher, Jeffrey T
>Cc: davem@davemloft.net; Tantilov, Emil S;
>netdev@vger.kernel.org; gospo@redhat.com;
>sassmann@redhat.com
>Subject: Re: [net-next 07/15] ixgbe: fix SFF data dumps of
>SFP+ modules
>
>On Mon, 2013-07-29 at 05:52 -0700, Jeff Kirsher wrote:
>> From: Emil Tantilov <emil.s.tantilov@intel.com>
>>
>> This patch fixes several issues with the previous
>implementation of the
>> SFF data dump of SFP+ modules:
>>
>> - removed the __IXGBE_READ_I2C flag - I2C access locking
>is handled in the
>> HW specific routines
>>
>> - fixed the read loop to read data from ee->offset to ee-
>>len
>>
>> - the reads fail if __IXGBE_IN_SFP_INIT is set in the
>process - this is
>> needed because on some HW I2C operations can take long
>time and disrupt
>> the SFP and link detection process
>>
>> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
>> Reported-by: Ben Hutchings <bhutchings@solarflare.com>
>> Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
>> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>[...]
>> @@ -2969,48 +2955,25 @@ static int
>ixgbe_get_module_eeprom(struct net_device *dev,
>> int i = 0;
>> int ret_val = 0;
>>
>> - /* ixgbe_get_module_info is called before this function
>in all
>> - * cases, so we do not need any checks we already do
>above,
>> - * and can trust ee->len to be a known value.
>> - */
>> + if (ee->len == 0)
>> + return -EINVAL;
>>
>> - while (test_bit(__IXGBE_IN_SFP_INIT, &adapter->state))
>> - msleep(100);
>> - set_bit(__IXGBE_READ_I2C, &adapter->state);
>> + for (i = ee->offset; i < ee->len; i++) {
>
> i < ee->offset + ee->len
>
>> + /* I2C reads can take long time */
>> + if (test_bit(__IXGBE_IN_SFP_INIT, &adapter-
>>state))
>> + return -EBUSY;
>>
>> - /* Read the first block, SFF-8079 */
>> - for (i = 0; i < ETH_MODULE_SFF_8079_LEN; i++) {
>> - status = hw->phy.ops.read_i2c_eeprom(hw, i,
>&databyte);
>> - if (status != 0) {
>> - /* Error occured while reading module */
>> + if (i < ETH_MODULE_SFF_8079_LEN)
>> + status = hw->phy.ops.read_i2c_eeprom(hw, i,
>&databyte);
>> + else
>> + status = hw->phy.ops.read_i2c_sff8472(hw, i,
>&databyte);
>[...]
> i -
>ETH_MODULE_SFF_8079_LEN ?
>
>But this works anyway because the address is truncated to
>u8.
Actually as is the loop will only work when offset is 0, so it looks like it should be:
for (i = ee->offset; i < ee->offset + ee->len; i++) {
Unless you had something else in mind.
I will submit a fix for this.
Thanks,
Emil
>
>Ben.
^ permalink raw reply
* [PATCH v3 4/4] USBNET: ax88179_178a: enable tso if usb host supports sg dma
From: Ming Lei @ 2013-08-06 0:52 UTC (permalink / raw)
To: David S. Miller, Greg Kroah-Hartman
Cc: Oliver Neukum, Sarah Sharp, netdev, linux-usb, Ming Lei,
Eric Dumazet, Ben Hutchings, Grant Grundler, Alan Stern,
Freddy Xin
In-Reply-To: <1375750370-18194-1-git-send-email-ming.lei@canonical.com>
This patch enables 'can_dma_sg' flag for ax88179_178a device
if the attached host controller supports building packet from
discontinuous buffers(DMA SG is possible), so TSO can be enabled
and skb fragment buffers can be passed to usb stack via urb->sg
directly.
With the patch, system CPU utilization decreased ~50% and throughput
increased by ~10% when doing iperf client test on one ARM A15 dual
core board.
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Grant Grundler <grundler@google.com>
Cc: Oliver Neukum <oneukum@suse.de>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Freddy Xin <freddy@asix.com.tw>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
---
drivers/net/usb/ax88179_178a.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/drivers/net/usb/ax88179_178a.c b/drivers/net/usb/ax88179_178a.c
index fb0caa2..7a96326 100644
--- a/drivers/net/usb/ax88179_178a.c
+++ b/drivers/net/usb/ax88179_178a.c
@@ -1031,12 +1031,20 @@ static int ax88179_bind(struct usbnet *dev, struct usb_interface *intf)
dev->mii.phy_id = 0x03;
dev->mii.supports_gmii = 1;
+ if (usb_device_no_sg_constraint(dev->udev))
+ dev->can_dma_sg = 1;
+
dev->net->features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
NETIF_F_RXCSUM;
dev->net->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
NETIF_F_RXCSUM;
+ if (dev->can_dma_sg) {
+ dev->net->features |= NETIF_F_SG | NETIF_F_TSO;
+ dev->net->hw_features |= NETIF_F_SG | NETIF_F_TSO;
+ }
+
/* Enable checksum offload */
*tmp = AX_RXCOE_IP | AX_RXCOE_TCP | AX_RXCOE_UDP |
AX_RXCOE_TCPV6 | AX_RXCOE_UDPV6;
@@ -1310,6 +1318,10 @@ static int ax88179_reset(struct usbnet *dev)
dev->net->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
NETIF_F_RXCSUM;
+ if (dev->can_dma_sg) {
+ dev->net->features |= NETIF_F_SG | NETIF_F_TSO;
+ dev->net->hw_features |= NETIF_F_SG | NETIF_F_TSO;
+ }
/* Enable checksum offload */
*tmp = AX_RXCOE_IP | AX_RXCOE_TCP | AX_RXCOE_UDP |
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 3/4] USBNET: support DMA SG
From: Ming Lei @ 2013-08-06 0:52 UTC (permalink / raw)
To: David S. Miller, Greg Kroah-Hartman
Cc: Oliver Neukum, Sarah Sharp, netdev, linux-usb, Ming Lei,
Eric Dumazet, Ben Hutchings, Grant Grundler, Freddy Xin,
Alan Stern
In-Reply-To: <1375750370-18194-1-git-send-email-ming.lei@canonical.com>
This patch introduces support of DMA SG if the USB host controller
which usbnet device is attached to is capable of building packet from
discontinuous buffers.
The patch supports passing the skb fragment buffers to usb stack directly
via urb->sg.
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Grant Grundler <grundler@google.com>
Cc: Freddy Xin <freddy@asix.com.tw>
Cc: Oliver Neukum <oneukum@suse.de>
Cc: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
---
drivers/net/usb/usbnet.c | 45 +++++++++++++++++++++++++++++++++++++++++---
include/linux/usb/usbnet.h | 1 +
2 files changed, 43 insertions(+), 3 deletions(-)
diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c
index e4811d7..534b60b 100644
--- a/drivers/net/usb/usbnet.c
+++ b/drivers/net/usb/usbnet.c
@@ -1232,6 +1232,37 @@ EXPORT_SYMBOL_GPL(usbnet_tx_timeout);
/*-------------------------------------------------------------------------*/
+static int build_dma_sg(const struct sk_buff *skb, struct urb *urb)
+{
+ unsigned num_sgs, total_len = 0;
+ int i, s = 0;
+
+ num_sgs = skb_shinfo(skb)->nr_frags + 1;
+ if (num_sgs == 1)
+ return 0;
+
+ urb->sg = kmalloc(num_sgs * sizeof(struct scatterlist), GFP_ATOMIC);
+ if (!urb->sg)
+ return -ENOMEM;
+
+ urb->num_sgs = num_sgs;
+ sg_init_table(urb->sg, urb->num_sgs);
+
+ sg_set_buf(&urb->sg[s++], skb->data, skb_headlen(skb));
+ total_len += skb_headlen(skb);
+
+ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
+ struct skb_frag_struct *f = &skb_shinfo(skb)->frags[i];
+
+ total_len += skb_frag_size(f);
+ sg_set_page(&urb->sg[i + s], f->page.p, f->size,
+ f->page_offset);
+ }
+ urb->transfer_buffer_length = total_len;
+
+ return 1;
+}
+
netdev_tx_t usbnet_start_xmit (struct sk_buff *skb,
struct net_device *net)
{
@@ -1258,7 +1289,6 @@ netdev_tx_t usbnet_start_xmit (struct sk_buff *skb,
goto drop;
}
}
- length = skb->len;
if (!(urb = usb_alloc_urb (0, GFP_ATOMIC))) {
netif_dbg(dev, tx_err, dev->net, "no urb\n");
@@ -1268,10 +1298,14 @@ netdev_tx_t usbnet_start_xmit (struct sk_buff *skb,
entry = (struct skb_data *) skb->cb;
entry->urb = urb;
entry->dev = dev;
- entry->length = length;
usb_fill_bulk_urb (urb, dev->udev, dev->out,
skb->data, skb->len, tx_complete, skb);
+ if (dev->can_dma_sg) {
+ if (build_dma_sg(skb, urb) < 0)
+ goto drop;
+ }
+ entry->length = length = urb->transfer_buffer_length;
/* don't assume the hardware handles USB_ZERO_PACKET
* NOTE: strictly conforming cdc-ether devices should expect
@@ -1340,7 +1374,10 @@ drop:
not_drop:
if (skb)
dev_kfree_skb_any (skb);
- usb_free_urb (urb);
+ if (urb) {
+ kfree(urb->sg);
+ usb_free_urb(urb);
+ }
} else
netif_dbg(dev, tx_queued, dev->net,
"> tx, len %d, type 0x%x\n", length, skb->protocol);
@@ -1391,6 +1428,7 @@ static void usbnet_bh (unsigned long param)
rx_process (dev, skb);
continue;
case tx_done:
+ kfree(entry->urb->sg);
case rx_cleanup:
usb_free_urb (entry->urb);
dev_kfree_skb (skb);
@@ -1727,6 +1765,7 @@ int usbnet_resume (struct usb_interface *intf)
retval = usb_submit_urb(res, GFP_ATOMIC);
if (retval < 0) {
dev_kfree_skb_any(skb);
+ kfree(res->sg);
usb_free_urb(res);
usb_autopm_put_interface_async(dev->intf);
} else {
diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h
index 8fbc008..9cb2fe8 100644
--- a/include/linux/usb/usbnet.h
+++ b/include/linux/usb/usbnet.h
@@ -35,6 +35,7 @@ struct usbnet {
unsigned char suspend_count;
unsigned char pkt_cnt, pkt_err;
unsigned short rx_qlen, tx_qlen;
+ unsigned can_dma_sg:1;
/* i/o info: pipes etc */
unsigned in, out;
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 2/4] USB: XHCI: mark no_sg_constraint
From: Ming Lei @ 2013-08-06 0:52 UTC (permalink / raw)
To: David S. Miller, Greg Kroah-Hartman
Cc: Oliver Neukum, Sarah Sharp, netdev, linux-usb, Ming Lei,
Alan Stern
In-Reply-To: <1375750370-18194-1-git-send-email-ming.lei@canonical.com>
This patch marks all xHCI controllers as no_sg_constraint
since xHCI supports building packet from discontinuous buffers.
Cc: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
---
drivers/usb/host/xhci.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
index 2c49f00..6e2ac57 100644
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -4841,6 +4841,10 @@ int xhci_gen_setup(struct usb_hcd *hcd, xhci_get_quirks_t get_quirks)
/* Accept arbitrarily long scatter-gather lists */
hcd->self.sg_tablesize = ~0;
+
+ /* support to build packet from discontinuous buffers */
+ hcd->self.no_sg_constraint = 1;
+
/* XHCI controllers don't stop the ep queue on short packets :| */
hcd->self.no_stop_on_short = 1;
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 1/4] USB: introduce usb_device_no_sg_constraint() helper
From: Ming Lei @ 2013-08-06 0:52 UTC (permalink / raw)
To: David S. Miller, Greg Kroah-Hartman
Cc: Oliver Neukum, Sarah Sharp, netdev, linux-usb, Ming Lei
In-Reply-To: <1375750370-18194-1-git-send-email-ming.lei@canonical.com>
Some host controllers(such as xHCI) can support building
packet from discontinuous buffers, so introduce one flag
and helper for this kind of host controllers, then the
feature can help some applications(such as usbnet) by
supporting arbitrary length of sg buffers.
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
---
drivers/usb/core/urb.c | 3 ++-
include/linux/usb.h | 8 +++++++-
2 files changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/core/urb.c b/drivers/usb/core/urb.c
index e75115a..c77ec78 100644
--- a/drivers/usb/core/urb.c
+++ b/drivers/usb/core/urb.c
@@ -414,7 +414,8 @@ int usb_submit_urb(struct urb *urb, gfp_t mem_flags)
urb->iso_frame_desc[n].status = -EXDEV;
urb->iso_frame_desc[n].actual_length = 0;
}
- } else if (dev->speed != USB_SPEED_WIRELESS && urb->num_sgs) {
+ } else if (urb->num_sgs && !urb->dev->bus->no_sg_constraint &&
+ dev->speed != USB_SPEED_WIRELESS) {
struct scatterlist *sg;
int i;
diff --git a/include/linux/usb.h b/include/linux/usb.h
index 84f14e2..bbd2c8d 100644
--- a/include/linux/usb.h
+++ b/include/linux/usb.h
@@ -337,6 +337,7 @@ struct usb_bus {
* the ep queue on a short transfer
* with the URB_SHORT_NOT_OK flag set.
*/
+ unsigned no_sg_constraint:1; /* no sg constraint */
unsigned sg_tablesize; /* 0 or largest number of sg list entries */
int devnum_next; /* Next open device number in
@@ -684,6 +685,11 @@ static inline bool usb_device_supports_ltm(struct usb_device *udev)
return udev->bos->ss_cap->bmAttributes & USB_LTM_SUPPORT;
}
+static inline bool usb_device_no_sg_constraint(struct usb_device *udev)
+{
+ return udev && udev->bus && udev->bus->no_sg_constraint;
+}
+
/*-------------------------------------------------------------------------*/
@@ -1249,7 +1255,7 @@ typedef void (*usb_complete_t)(struct urb *);
* transfer_buffer.
* @sg: scatter gather buffer list, the buffer size of each element in
* the list (except the last) must be divisible by the endpoint's
- * max packet size
+ * max packet size if no_sg_constraint isn't set in 'struct usb_bus'
* @num_mapped_sgs: (internal) number of mapped sg entries
* @num_sgs: number of entries in the sg list
* @transfer_buffer_length: How big is transfer_buffer. The transfer may
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 0/4] USB & USBNET: loose SG check and support usbnet DMA SG
From: Ming Lei @ 2013-08-06 0:52 UTC (permalink / raw)
To: David S. Miller, Greg Kroah-Hartman
Cc: Oliver Neukum, Sarah Sharp, netdev, linux-usb
Hi,
This patchset allows drivers to pass sg buffers which size can't be divided
by max packet size of endpoint if the host controllers(such ax xHCI) support
this kind of sg buffers.
Previously we added check[1] on the situation and don't allow these sg buffers
passed to USB HCD, looks the check is too strict to make use of new feature of
new hardware(xHCI) for some applications(such as network stack) which can't
provide this kind of sg buffers to usbnet driver, so the patch looses the check
in case that the host controller supports it.
Patch 3/4 implements DMA SG on usbnet driver, and patch 4/4 uses it on ax88179_178a
USB3 NIC for supporting TSO, so both CPU utilization and tx throughput can be
improved with TSO and DMA SG in case of the USB NIC is attached to xHCI controller.
This patchset depends on both net-next and usb-next tree, so hope David and Greg
to figure out one elegent way to merge it.
[1], http://git.kernel.org/cgit/linux/kernel/git/gregkh/usb.git/commit/?h=usb-next&id=10e232c597ac757e7f8600649f7e872e86de190f
V3:
- save 3 lines code for usb_device_no_sg_constraint() as suggested by Alan
- fix urb->sg leak in xmit failure path
V2:
- add missed kfree(urb->sg) in 3/4
- rename no_sg_limit as no_sg_constraint as suggested by Alan
V1:
- introduce and apply usb_device_no_sg_limit() helper as suggested by Greg
- simplify patch 4/4 against Eric Dumazet's patch(ax88179_178a: avoid copy of tx tcp packets)
- don't pass usbnet header as sg buffer
drivers/net/usb/ax88179_178a.c | 12 +++++++++++
drivers/net/usb/usbnet.c | 45 +++++++++++++++++++++++++++++++++++++---
drivers/usb/core/urb.c | 3 ++-
drivers/usb/host/xhci.c | 4 ++++
include/linux/usb.h | 8 ++++++-
include/linux/usb/usbnet.h | 1 +
6 files changed, 68 insertions(+), 5 deletions(-)
Thanks,
--
Ming Lei
^ permalink raw reply
* [PATCH] net/ipv4: fix the conditions of entering TCP_CA_Disorder state
From: Dong Fang @ 2013-08-06 0:45 UTC (permalink / raw)
To: davem, kuznet, jmorris, yoshfuji, kaber; +Cc: netdev, linux-kernel, Dong Fang
if have some packets loss by network, the kernel can't reach here, we can see
the tcp_time_to_recover() function:
static bool tcp_time_to_recover(struct sock *sk, int flag)
{
struct tcp_sock *tp = tcp_sk(sk);
__u32 packets_out;
/* Trick#1: The loss is proven. */
if (tp->lost_out)
return true;
//...
}
when it return true, the following condition will be failed:
//...
if (!tcp_time_to_recover(sk, flag)) {
tcp_try_to_open(sk, flag, prior_unsacked);
return;
}
//...
Signed-off-by: Dong Fang <yp.fangdong@gmail.com>
---
net/ipv4/tcp_input.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 28af45a..8ab31c1 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -2449,7 +2449,7 @@ static void tcp_try_keep_open(struct sock *sk)
struct tcp_sock *tp = tcp_sk(sk);
int state = TCP_CA_Open;
- if (tcp_left_out(tp) || tcp_any_retrans_done(sk))
+ if (tp->sacked_out || tcp_any_retrans_done(sk))
state = TCP_CA_Disorder;
if (inet_csk(sk)->icsk_ca_state != state) {
--
1.7.1
^ permalink raw reply related
* [PATCH] bridge: correct the comment for file br_sysfs_br.c
From: Wang Sheng-Hui @ 2013-08-06 0:44 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge, netdev
br_sysfs_if.c is for sysfs attributes of bridge ports, while br_sysfs_br.c
is for sysfs attributes of bridge itself. Correct the comment here.
Signed-off-by: Wang Sheng-Hui <shhuiw@gmail.com>
---
net/bridge/br_sysfs_br.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/bridge/br_sysfs_br.c b/net/bridge/br_sysfs_br.c
index 394bb96..3b9637f 100644
--- a/net/bridge/br_sysfs_br.c
+++ b/net/bridge/br_sysfs_br.c
@@ -1,5 +1,5 @@
/*
- * Sysfs attributes of bridge ports
+ * Sysfs attributes of bridge
* Linux ethernet bridge
*
* Authors:
--
1.7.10.4
^ permalink raw reply related
* Re: [net-next,1/3] bonding: fix vlan 0 addition and removal
From: Veaceslav Falico @ 2013-08-06 0:37 UTC (permalink / raw)
To: David Miller; +Cc: nikolay, netdev, fubar, andy, kaber
In-Reply-To: <20130805.160822.47868741818443064.davem@davemloft.net>
On Mon, Aug 05, 2013 at 04:08:22PM -0700, David Miller wrote:
>From: Veaceslav Falico <vfalico@redhat.com>
>Date: Mon, 5 Aug 2013 23:51:26 +0200
>
>> @@ -69,7 +69,6 @@
>> #include <net/arp.h>
>> #include <linux/mii.h>
>> #include <linux/ethtool.h>
>> -#include <linux/if_vlan.h>
>> #include <linux/if_bonding.h>
>> #include <linux/jiffies.h>
>> #include <linux/preempt.h>
>> @@ -1976,7 +1975,7 @@ static int __bond_release_one(struct net_device
>> *bond_dev,
>> bond_set_carrier(bond);
>> eth_hw_addr_random(bond_dev);
>> - if (bond_vlan_used(bond)) {
>> + if (vlan_uses_dev(bond_dev)) {
>
>If you're adding a use of vlan_uses_dev(), you should retain the
>if_vlan.h include, not remove it.
I've added the usage of vlan_uses_dev_rcu() to bonding.h, so bonding.h
already includes linux/if_vlan.h, and thus any other bonding file doesn't
need it, cause it always includes bonding.h.
Sorry, my bad, should have added it to the commit message.
^ permalink raw reply
* Re: [PATCH] bridge: don't try to update timers in case of broken MLD queries
From: Linus Lüssing @ 2013-08-06 0:32 UTC (permalink / raw)
To: Stephen Hemminger
Cc: Paul Bolle, Cong Wang, netdev, bridge, linux-kernel, Adam Baker,
linus.luessing, David S. Miller, Herbert Xu
In-Reply-To: <20130805154222.58361734@nehalam.linuxnetplumber.net>
On Mon, Aug 05, 2013 at 03:42:22PM -0700, Stephen Hemminger wrote:
> On Tue, 6 Aug 2013 00:32:05 +0200
> Linus Lüssing <linus.luessing@web.de> wrote:
>
> > Currently we are reading an uninitialized value for the max_delay
> > variable when snooping an MLD query message of invalid length and would
> > update our timers with that.
> >
> > Fixing this by simply ignoring such broken MLD queries (just like we do
> > for IGMP already).
> >
> > This is a regression introduced by:
> > "bridge: disable snooping if there is no querier" (b00589af3b04)
> >
> > Reported-by: Paul Bolle <pebolle@tiscali.nl>
> > Signed-off-by: Linus Lüssing <linus.luessing@web.de>
> > ---
> > net/bridge/br_multicast.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
> > index 61c5e81..08e576a 100644
> > --- a/net/bridge/br_multicast.c
> > +++ b/net/bridge/br_multicast.c
> > @@ -1195,7 +1195,7 @@ static int br_ip6_multicast_query(struct net_bridge *br,
> > max_delay = msecs_to_jiffies(ntohs(mld->mld_maxdelay));
> > if (max_delay)
> > group = &mld->mld_mca;
> > - } else if (skb->len >= sizeof(*mld2q)) {
> > + } else {
> > if (!pskb_may_pull(skb, sizeof(*mld2q))) {
> > err = -EINVAL;
> > goto out;
>
> Why not use else if here, other than that looks great.
Because it isn't really necessary, it is basically included
in the pskb_may_pull() already, just like it is in the according IGMP
code path.
And I thought it'd be nicer to handle it the same way as in the
IGMP code path to avoid diverging too much.
Cheers, Linus
^ permalink raw reply
* Re: BUG cxgb3: Check and handle the dma mapping errors
From: Alexey Kardashevskiy @ 2013-08-06 0:15 UTC (permalink / raw)
To: Jay Fenlason
Cc: Santosh Rastapur, Divy Le Ray, David S. Miller, netdev,
Linux Kernel Mailing List
In-Reply-To: <20130805184107.GA2998@redhat.com>
On 08/06/2013 04:41 AM, Jay Fenlason wrote:
> On Mon, Aug 05, 2013 at 12:59:04PM +1000, Alexey Kardashevskiy wrote:
>> Hi!
>>
>> Recently I started getting multiple errors like this:
>>
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> cxgb3 0006:01:00.0: iommu_alloc failed, tbl c000000003067980 vaddr
>> c000001fbdaaa882 npages 1
>> ... and so on
>>
>> This is all happening on a PPC64 "powernv" platform machine. To trigger the
>> error state, it is enough to _flood_ ping CXGB3 card from another machine
>> (which has Emulex 10Gb NIC + Cisco switch). Just do "ping -f 172.20.1.2"
>> and wait 10-15 seconds.
>>
>>
>> The messages are coming from arch/powerpc/kernel/iommu.c and basically
>> mean that the driver requested more pages than the DMA window has which is
>> normally 1GB (there could be another possible source of errors -
>> ppc_md.tce_build callback - but on powernv platform it always succeeds).
>>
>>
>> The patch after which it broke is:
>> commit f83331bab149e29fa2c49cf102c0cd8c3f1ce9f9
>> Author: Santosh Rastapur <santosh@chelsio.com>
>> Date: Tue May 21 04:21:29 2013 +0000
>> cxgb3: Check and handle the dma mapping errors
>>
>> Any quick ideas? Thanks!
>
> That patch adds error checking to detect failed dma mapping requests.
> Before it, the code always assumed that dma mapping requests succeded,
> whether they actually do or not, so the fact that the older kernel
> does not log errors only means that the failures are being ignored,
> and any appearance of working is through pure luck. The machine could
> have just crashed at that point.
>From what I see, the patch adds map_skb() function which is called in two
new places, so the patch does not just mechanically replace
skb_frag_dma_map() to map_skb() or something like that.
> What is the observed behavior of the system by the machine initiating
> the ping flood? Do the older and newer kernels differ in the
> percentage of pings that do not receive replies?
The other machine stops receiving replies. It is using different adapter,
not Chelsio and the kernel version does not really matter.
> O the newer kernel,
> when the mapping errors are detected, the packet that it is trying to
> transmit is dropped, but I'm not at all sure what happens on the older
> kernel after the dma mapping fails. As I mentioned earlier, I'm
> surprised it does not crash. Perhaps the folks from Chelsio have a
> better idea what happens after a dma mapping error is ignored?
Any kernel cannot avoid platform's iommu_alloc() on ppc64/powernv so if
there was a problem, we would have seen messages (and yes, kernel would
have crashed).
--
Alexey
^ permalink raw reply
* [PATCH] tcp: cubic: fix overflow error in bictcp_update()
From: Eric Dumazet @ 2013-08-06 0:10 UTC (permalink / raw)
To: David Miller
Cc: netdev, Neal Cardwell, Van Jacobson, Stephen Hemminger,
Yuchung Cheng
From: Eric Dumazet <edumazet@google.com>
commit 17a6e9f1aa9 ("tcp_cubic: fix clock dependency") added an
overflow error in bictcp_update() in following code :
/* change the unit from HZ to bictcp_HZ */
t = ((tcp_time_stamp + msecs_to_jiffies(ca->delay_min>>3) -
ca->epoch_start) << BICTCP_HZ) / HZ;
Because msecs_to_jiffies() being unsigned long, compiler does
implicit type promotion.
We really want to constrain (tcp_time_stamp - ca->epoch_start)
to a signed 32bit value, or else 't' has unexpected high values.
This bugs triggers an increase of retransmit rates ~24 days after
boot [1], as the high order bit of tcp_time_stamp flips.
[1] for hosts with HZ=1000
Big thanks to Van Jacobson for spotting this problem.
Diagnosed-by: Van Jacobson <vanj@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
---
net/ipv4/tcp_cubic.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index a9077f4..b6b591f 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -206,8 +206,8 @@ static u32 cubic_root(u64 a)
*/
static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
{
- u64 offs;
- u32 delta, t, bic_target, max_cnt;
+ u32 delta, bic_target, max_cnt;
+ u64 offs, t;
ca->ack_cnt++; /* count the number of ACKs */
@@ -250,9 +250,11 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
* if the cwnd < 1 million packets !!!
*/
+ t = (s32)(tcp_time_stamp - ca->epoch_start);
+ t += msecs_to_jiffies(ca->delay_min >> 3);
/* change the unit from HZ to bictcp_HZ */
- t = ((tcp_time_stamp + msecs_to_jiffies(ca->delay_min>>3)
- - ca->epoch_start) << BICTCP_HZ) / HZ;
+ t <<= BICTCP_HZ;
+ do_div(t, HZ);
if (t < ca->bic_K) /* t - K */
offs = ca->bic_K - t;
^ permalink raw reply related
* Re: [PATCH net v4 2/6] bnx2x: protect different statistics flows
From: David Miller @ 2013-08-05 23:26 UTC (permalink / raw)
To: dmitry; +Cc: netdev, eilong, ariele
In-Reply-To: <1375738544-8695-3-git-send-email-dmitry@broadcom.com>
From: "Dmitry Kravkov" <dmitry@broadcom.com>
Date: Tue, 6 Aug 2013 00:35:40 +0300
> Add locking to protect different statistics flows from
> running simultaneously.
> This in order to serialize statistics requests sent to FW,
> otherwise two outstanding queries may cause FW assert.
>
> Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
> Signed-off-by: Ariel Elior <ariele@broadcom.com>
> Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Please do not reinvent the wheel, use a standard semaphore
and down_timeout().
Anytime you feel that you have to implement your own fancy locking
primitives, you are doing it wrong.
^ permalink raw reply
* Re: [PATCH net-next] bonding: fix send_peer_notif leekage on rtnl lock congestion
From: David Miller @ 2013-08-05 23:12 UTC (permalink / raw)
To: vfalico; +Cc: netdev, fubar, andy
In-Reply-To: <1375732053-829-1-git-send-email-vfalico@redhat.com>
From: Veaceslav Falico <vfalico@redhat.com>
Date: Mon, 5 Aug 2013 21:47:33 +0200
> @@ -2999,7 +3002,10 @@ void bond_activebackup_arp_mon(struct work_struct *work)
> if (!rtnl_trylock()) {
> read_lock(&bond->lock);
> delta_in_ticks = 1;
> - should_notify_peers = false;
> + if (should_notify_peers) {
> + bond->send_peer_notif++;
I doubt this increment to a shared datastructure is safe with
the locks you hold here.
You don't hold RTNL and you only have bond->lock as a reader.
^ permalink raw reply
* Re: [net-next,3/3] bonding: unwind on bond_add_vlan add failure
From: David Miller @ 2013-08-05 23:09 UTC (permalink / raw)
To: vfalico; +Cc: nikolay, netdev, fubar, andy
In-Reply-To: <20130805215921.GA14299@redhat.com>
From: Veaceslav Falico <vfalico@redhat.com>
Date: Mon, 5 Aug 2013 23:59:21 +0200
> On Mon, Aug 05, 2013 at 03:28:24PM +0200, nikolay@redhat.com wrote:
>>From: Nikolay Aleksandrov <Nikolay Aleksandrov nikolay@redhat.com>
>>
>>In case of bond_add_vlan() failure currently we'll have the vlan's
>>refcnt bumped up in all slaves, but it will never go down because it
>>failed to get added to the bond, so properly unwind the added vlan if
>>bond_add_vlan fails.
>>
>>Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com>
>
> In case patch 1/3 from series goes in... Otherwise I think it will be
> needed to patch bond_vlan_rx_add_vid().
I'm not happy with patch #1 and it needs more discussion between you
two.
^ permalink raw reply
* Re: [PATCH] lto, wan/sbni: Make inline assembler symbols visible and assembler global
From: David Miller @ 2013-08-05 23:09 UTC (permalink / raw)
To: andi; +Cc: netdev, ak
In-Reply-To: <20130805230535.GP19750@two.firstfloor.org>
From: Andi Kleen <andi@firstfloor.org>
Date: Tue, 6 Aug 2013 01:05:35 +0200
> On Mon, Aug 05, 2013 at 03:33:23PM -0700, David Miller wrote:
>> From: Andi Kleen <andi@firstfloor.org>
>> Date: Mon, 5 Aug 2013 15:18:14 -0700
>>
>> > From: Andi Kleen <ak@linux.intel.com>
>> >
>> > - Inline assembler defining C callable code has to be global
>> > - The function has to be visible
>> >
>> > Do this in wan/sbni
>> >
>> > Signed-off-by: Andi Kleen <ak@linux.intel.com>
>>
>> Please just make it use the standard lib/ CRC routines instead.
>
> The problem is I don't have this hardware, and I don't think
> I can do such a change without testing it.
I say you can, having assembly in here for something as fundamental
as this is pointless.
^ permalink raw reply
* Re: [net-next,1/3] bonding: fix vlan 0 addition and removal
From: David Miller @ 2013-08-05 23:08 UTC (permalink / raw)
To: vfalico; +Cc: nikolay, netdev, fubar, andy, kaber
In-Reply-To: <20130805215126.GB3859@redhat.com>
From: Veaceslav Falico <vfalico@redhat.com>
Date: Mon, 5 Aug 2013 23:51:26 +0200
> @@ -69,7 +69,6 @@
> #include <net/arp.h>
> #include <linux/mii.h>
> #include <linux/ethtool.h>
> -#include <linux/if_vlan.h>
> #include <linux/if_bonding.h>
> #include <linux/jiffies.h>
> #include <linux/preempt.h>
> @@ -1976,7 +1975,7 @@ static int __bond_release_one(struct net_device
> *bond_dev,
> bond_set_carrier(bond);
> eth_hw_addr_random(bond_dev);
> - if (bond_vlan_used(bond)) {
> + if (vlan_uses_dev(bond_dev)) {
If you're adding a use of vlan_uses_dev(), you should retain the
if_vlan.h include, not remove it.
^ permalink raw reply
* Re: [PATCH] lto, wan/sbni: Make inline assembler symbols visible and assembler global
From: Andi Kleen @ 2013-08-05 23:05 UTC (permalink / raw)
To: David Miller; +Cc: andi, netdev, ak
In-Reply-To: <20130805.153323.276242626385124232.davem@davemloft.net>
On Mon, Aug 05, 2013 at 03:33:23PM -0700, David Miller wrote:
> From: Andi Kleen <andi@firstfloor.org>
> Date: Mon, 5 Aug 2013 15:18:14 -0700
>
> > From: Andi Kleen <ak@linux.intel.com>
> >
> > - Inline assembler defining C callable code has to be global
> > - The function has to be visible
> >
> > Do this in wan/sbni
> >
> > Signed-off-by: Andi Kleen <ak@linux.intel.com>
>
> Please just make it use the standard lib/ CRC routines instead.
The problem is I don't have this hardware, and I don't think
I can do such a change without testing it.
FWIW my main interest here is to avoid this thing breaking
my allyesconfig build.
Could you please consider adding it?
If not I'll have to disable the driver in the LTO patchkit.
-Andi
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox