On Fri, Feb 15, 2013 at 08:58:52AM -0800, Ben Pfaff wrote:
> On Fri, Feb 15, 2013 at 06:55:55PM +0900, Simon Horman wrote:
> > When encoding MPLS information in ODP flow key data use encap()
> > for encap for frames. This seems to be in keeping with datapath/README.
> >
> > before: eth(...), eth_type(0x8847), mpls(...), encap(ip(...), tcp(...))
> > after: eth(...), eth_type(0x8847), mpls(...), ip(...), tcp(...)
>
> I think you got the before and after backward.
Thanks, I have fixed that below.
> We also need to update odp_flow_key_from_flow().
Thanks, I had that in the next patch in the series.
A combined patch with a revised commit message is below.
From: Simon Horman <ho...@verge.net.au>
odp: Use encap() for MPLS-encapsulated ODP flow key data
This patch has two components:
1. odp_flow_key_from_flow() is updated to extract encapsulated frame
information from MPLS flows. It does this by encoding the encapsualted
frame in an encap() in the resulting ODP key data.
A subsequent patch will introduce code to set flow->encap_dl_type
which will activate this change.
2. Update parse_l2_5_onward() to read an MPLS-encapsulated frame
from inside encap() in ODP key data. This consumes ODP key data
produced by odp_flow_key_from_flow().
before: eth(...), eth_type(0x8847), mpls(...), ip(...), tcp(...)
after: eth(...), eth_type(0x8847), mpls(...), encap(ip(...), tcp(...))
The use of encap() for MPLS-encapsulated frames was suggested by Ben Pfaff
and seems to be in keeping with datapath/README.
Signed-off-by: Simon Horman <ho...@verge.net.au>
---
v2.20 (draft)
* Squash "Update parse_l2_5_onward to use encap for MPLS" and
"odp_flow_key_from_flow: Allow of encapsulated frames" and
rename the result "odp: Use encap() for MPLS-encapsulated ODP flow key data".
v2.19
* As suggested by Ben Pfaff
- Remove encap_dl_type parameter from odp_flow_key_from_flow()
- Use encap for mpls in ODP flow key data.
before: eth(...), eth_type(0x8847), mpls(...), ip(...), tcp(...)
after: eth(...), eth_type(0x8847), mpls(...), encap(ip(...), tcp(...))
* Coding style fix for actions_allow_l3_extraction().
Use {} in conjunction with if statements.
* Enhance actions_allow_l3_extraction to return when the first MPLS_PUSH
action with a non-MPLS ether type is found rather than when the first
MPLS_PUSH action is found.
* Remove actions_allow_l3_extraction() and obtain encap_dl_type
from xlate_actions() instead.
* Removed non-odp_flow_key_from_flow portions.
These are now present in ofproto: Allow richer matches based on actions().
* Rename patch from
"actions: Allow secondary decoding of a flow" to
"actions: odp_flow_key_from_flow: Allow of encapsulated frames"
v2.18
* Update odp_flow_key_from_flow() to use dl_type_is_ip_any() wrapper.
As odp_flow_key_from_flow() is paramatised over encap_dl_type
this seems to be a good approach.
v2.17
* Rebase
v2.13 - v2.16
* No change
v2.12
* As suggested by Jarno Rajahalme
- Use flow->encap_dl_type instead of of obtaining the value
by scanning the actions.
v2.11
* First post
---
lib/odp-util.c | 85 ++++++++++++++++++++++++++++++++-----------------
tests/ofproto-dpif.at | 4 +--
2 files changed, 58 insertions(+), 31 deletions(-)
diff --git a/lib/odp-util.c b/lib/odp-util.c
index f3f66b7..a6002db 100644
--- a/lib/odp-util.c
+++ b/lib/odp-util.c
@@ -1506,7 +1506,8 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
uint32_t odp_in_port)
{
struct ovs_key_ethernet *eth_key;
- size_t encap;
+ size_t encap_vlan = 0, encap_mpls = 0;
+ ovs_be16 dl_type = flow_innermost_dl_type(flow);
if (flow->skb_priority) {
nl_msg_put_u32(buf, OVS_KEY_ATTR_PRIORITY, flow->skb_priority);
@@ -1534,12 +1535,10 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
if (flow->vlan_tci != htons(0) || flow->dl_type == htons(ETH_TYPE_VLAN)) {
nl_msg_put_be16(buf, OVS_KEY_ATTR_ETHERTYPE, htons(ETH_TYPE_VLAN));
nl_msg_put_be16(buf, OVS_KEY_ATTR_VLAN, flow->vlan_tci);
- encap = nl_msg_start_nested(buf, OVS_KEY_ATTR_ENCAP);
+ encap_vlan = nl_msg_start_nested(buf, OVS_KEY_ATTR_ENCAP);
if (flow->vlan_tci == htons(0)) {
goto unencap;
}
- } else {
- encap = 0;
}
if (ntohs(flow->dl_type) < ETH_TYPE_MIN) {
@@ -1548,7 +1547,18 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
nl_msg_put_be16(buf, OVS_KEY_ATTR_ETHERTYPE, flow->dl_type);
- if (flow->dl_type == htons(ETH_TYPE_IP)) {
+ if (flow->mpls_depth) {
+ struct ovs_key_mpls *mpls_key;
+
+ mpls_key = nl_msg_put_unspec_uninit(buf, OVS_KEY_ATTR_MPLS,
+ sizeof *mpls_key);
+ mpls_key->mpls_top_lse = flow->mpls_lse;
+ if (flow->encap_dl_type) {
+ encap_mpls = nl_msg_start_nested(buf, OVS_KEY_ATTR_ENCAP);
+ }
+ }
+
+ if (dl_type == htons(ETH_TYPE_IP)) {
struct ovs_key_ipv4 *ipv4_key;
ipv4_key = nl_msg_put_unspec_uninit(buf, OVS_KEY_ATTR_IPV4,
@@ -1559,7 +1569,7 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
ipv4_key->ipv4_tos = flow->nw_tos;
ipv4_key->ipv4_ttl = flow->nw_ttl;
ipv4_key->ipv4_frag = ovs_to_odp_frag(flow->nw_frag);
- } else if (flow->dl_type == htons(ETH_TYPE_IPV6)) {
+ } else if (dl_type == htons(ETH_TYPE_IPV6)) {
struct ovs_key_ipv6 *ipv6_key;
ipv6_key = nl_msg_put_unspec_uninit(buf, OVS_KEY_ATTR_IPV6,
@@ -1571,8 +1581,8 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
ipv6_key->ipv6_tclass = flow->nw_tos;
ipv6_key->ipv6_hlimit = flow->nw_ttl;
ipv6_key->ipv6_frag = ovs_to_odp_frag(flow->nw_frag);
- } else if (flow->dl_type == htons(ETH_TYPE_ARP) ||
- flow->dl_type == htons(ETH_TYPE_RARP)) {
+ } else if (dl_type == htons(ETH_TYPE_ARP) ||
+ dl_type == htons(ETH_TYPE_RARP)) {
struct ovs_key_arp *arp_key;
arp_key = nl_msg_put_unspec_uninit(buf, OVS_KEY_ATTR_ARP,
@@ -1585,15 +1595,7 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
memcpy(arp_key->arp_tha, flow->arp_tha, ETH_ADDR_LEN);
}
- if (flow->mpls_depth) {
- struct ovs_key_mpls *mpls_key;
-
- mpls_key = nl_msg_put_unspec_uninit(buf, OVS_KEY_ATTR_MPLS,
- sizeof *mpls_key);
- mpls_key->mpls_top_lse = flow->mpls_lse;
- }
-
- if (is_ip_any(flow) && !(flow->nw_frag & FLOW_NW_FRAG_LATER)) {
+ if (dl_type_is_ip_any(dl_type) && !(flow->nw_frag & FLOW_NW_FRAG_LATER)) {
if (flow->nw_proto == IPPROTO_TCP) {
struct ovs_key_tcp *tcp_key;
@@ -1608,7 +1610,7 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
sizeof *udp_key);
udp_key->udp_src = flow->tp_src;
udp_key->udp_dst = flow->tp_dst;
- } else if (flow->dl_type == htons(ETH_TYPE_IP)
+ } else if (dl_type == htons(ETH_TYPE_IP)
&& flow->nw_proto == IPPROTO_ICMP) {
struct ovs_key_icmp *icmp_key;
@@ -1616,7 +1618,7 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
sizeof *icmp_key);
icmp_key->icmp_type = ntohs(flow->tp_src);
icmp_key->icmp_code = ntohs(flow->tp_dst);
- } else if (flow->dl_type == htons(ETH_TYPE_IPV6)
+ } else if (dl_type == htons(ETH_TYPE_IPV6)
&& flow->nw_proto == IPPROTO_ICMPV6) {
struct ovs_key_icmpv6 *icmpv6_key;
@@ -1640,8 +1642,11 @@ odp_flow_key_from_flow(struct ofpbuf *buf, const struct
flow *flow,
}
unencap:
- if (encap) {
- nl_msg_end_nested(buf, encap);
+ if (encap_mpls) {
+ nl_msg_end_nested(buf, encap_mpls);
+ }
+ if (encap_vlan) {
+ nl_msg_end_nested(buf, encap_vlan);
}
}
@@ -1801,10 +1806,14 @@ parse_l2_5_onward(const struct nlattr
*attrs[OVS_KEY_ATTR_MAX + 1],
const struct nlattr *key, size_t key_len)
{
static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 5);
+ enum odp_key_fitness fitness, mpls_fitness = 0;
ovs_be16 dl_type;
/* Parse MPLS label stack entry */
if (eth_type_mpls(flow->dl_type)) {
+ const struct nlattr *encap
+ = (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_ENCAP)
+ ? attrs[OVS_KEY_ATTR_ENCAP] : NULL);
/* Calculate fitness of outer attributes. */
expected_attrs |= (UINT64_C(1) << OVS_KEY_ATTR_MPLS);
@@ -1815,12 +1824,27 @@ parse_l2_5_onward(const struct nlattr
*attrs[OVS_KEY_ATTR_MAX + 1],
flow->mpls_lse = nl_attr_get_be32(attrs[OVS_KEY_ATTR_MPLS]);
flow->mpls_depth++;
- if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_IPV4)) {
- flow->encap_dl_type = htons(ETH_TYPE_IP);
- } else if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_IPV6)) {
- flow->encap_dl_type = htons(ETH_TYPE_IPV6);
- } else if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_ARP)) {
- flow->encap_dl_type = htons(ETH_TYPE_ARP);
+ /* Encap may not be present if nothing is known about the
+ * encapsulated frame */
+ if (encap) {
+ expected_attrs |= (UINT64_C(1) << OVS_KEY_ATTR_ENCAP);
+ mpls_fitness = check_expectations(present_attrs, out_of_range_attr,
+ expected_attrs, key, key_len);
+
+ /* Now parse the encapsulated attributes. */
+ if (!parse_flow_nlattrs(nl_attr_get(encap),
nl_attr_get_size(encap),
+ attrs, &present_attrs,
&out_of_range_attr)) {
+ return ODP_FIT_ERROR;
+ }
+ expected_attrs = 0;
+
+ if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_IPV4)) {
+ flow->encap_dl_type = htons(ETH_TYPE_IP);
+ } else if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_IPV6)) {
+ flow->encap_dl_type = htons(ETH_TYPE_IPV6);
+ } else if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_ARP)) {
+ flow->encap_dl_type = htons(ETH_TYPE_ARP);
+ }
}
}
@@ -1939,8 +1963,11 @@ parse_l2_5_onward(const struct nlattr
*attrs[OVS_KEY_ATTR_MAX + 1],
}
}
- return check_expectations(present_attrs, out_of_range_attr, expected_attrs,
- key, key_len);
+ fitness = check_expectations(present_attrs, out_of_range_attr,
+ expected_attrs, key, key_len);
+
+ /* The overall fitness is the worse of the outer and inner attributes. */
+ return MAX(fitness, mpls_fitness);
}
/* Parse 802.1Q header then encapsulated L3 attributes. */
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
index 8f11b82..15069bd 100644
--- a/tests/ofproto-dpif.at
+++ b/tests/ofproto-dpif.at
@@ -359,7 +359,7 @@ dnl Modified MPLS actions.
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=50:55:55:55:55:55,dst=50:54:00:00:00:07),eth_type(0x8847),mpls(label=100,tc=7,ttl=64,bos=1),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
+ ovs-appctl netdev-dummy/receive p1
'in_port(1),eth(src=50:55:55:55:55:55,dst=50:54:00:00:00:07),eth_type(0x8847),mpls(label=100,tc=7,ttl=64,bos=1),encap(ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
done
OVS_WAIT_UNTIL([ovs-appctl -t ovs-ofctl exit])
@@ -398,7 +398,7 @@ dnl Modified MPLS pop action.
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=60:66:66:66:66:66,dst=50:54:00:00:00:07),eth_type(0x8847),mpls(label=10,tc=3,ttl=100,bos=1),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)'
+ ovs-appctl netdev-dummy/receive p1
'in_port(1),eth(src=60:66:66:66:66:66,dst=50:54:00:00:00:07),eth_type(0x8847),mpls(label=10,tc=3,ttl=100,bos=1),encap(ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no))'
done
OVS_WAIT_UNTIL([ovs-appctl -t ovs-ofctl exit])
--
1.7.10.4
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
