* [dpdk-stable] [DPDK] net/ice: fix tunnel rule not recognize
[not found] <20190715093807.128101-1-qiming.yang@intel.com>
@ 2019-07-15 9:38 ` Qiming Yang
2019-07-15 9:38 ` [dpdk-stable] [PATCH v2] net/ice: fix flow validation fail issue Qiming Yang
1 sibling, 0 replies; 2+ messages in thread
From: Qiming Yang @ 2019-07-15 9:38 UTC (permalink / raw)
To: qabuild; +Cc: Qiming Yang, stable
In the past, to distinguish whether the input set is outer or inner
by check the item appear once or twice. But this way doesn't work when user
don't configure the outer input set. This patch fix the issue.
Fixes: d76116a4678f ("net/ice: add generic flow API")
Cc: stable@dpdk.org
Signed-off-by: Qiming Yang <qiming.yang@intel.com>
---
drivers/net/ice/ice_generic_flow.c | 80 ++++++++++++++++++--------------------
1 file changed, 38 insertions(+), 42 deletions(-)
diff --git a/drivers/net/ice/ice_generic_flow.c b/drivers/net/ice/ice_generic_flow.c
index e6a2c4b..05a1678 100644
--- a/drivers/net/ice/ice_generic_flow.c
+++ b/drivers/net/ice/ice_generic_flow.c
@@ -209,8 +209,7 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,
0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF };
uint64_t input_set = ICE_INSET_NONE;
- bool outer_ip = true;
- bool outer_l4 = true;
+ bool is_tunnel = false;
for (; item->type != RTE_FLOW_ITEM_TYPE_END; item++) {
if (item->last) {
@@ -259,27 +258,26 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
return 0;
}
- if (outer_ip) {
+ if (is_tunnel) {
if (ipv4_mask->hdr.src_addr == UINT32_MAX)
- input_set |= ICE_INSET_IPV4_SRC;
+ input_set |= ICE_INSET_TUN_IPV4_SRC;
if (ipv4_mask->hdr.dst_addr == UINT32_MAX)
- input_set |= ICE_INSET_IPV4_DST;
- if (ipv4_mask->hdr.type_of_service == UINT8_MAX)
- input_set |= ICE_INSET_IPV4_TOS;
+ input_set |= ICE_INSET_TUN_IPV4_DST;
if (ipv4_mask->hdr.time_to_live == UINT8_MAX)
- input_set |= ICE_INSET_IPV4_TTL;
+ input_set |= ICE_INSET_TUN_IPV4_TTL;
if (ipv4_mask->hdr.next_proto_id == UINT8_MAX)
- input_set |= ICE_INSET_IPV4_PROTO;
- outer_ip = false;
+ input_set |= ICE_INSET_TUN_IPV4_PROTO;
} else {
if (ipv4_mask->hdr.src_addr == UINT32_MAX)
- input_set |= ICE_INSET_TUN_IPV4_SRC;
+ input_set |= ICE_INSET_IPV4_SRC;
if (ipv4_mask->hdr.dst_addr == UINT32_MAX)
- input_set |= ICE_INSET_TUN_IPV4_DST;
+ input_set |= ICE_INSET_IPV4_DST;
if (ipv4_mask->hdr.time_to_live == UINT8_MAX)
- input_set |= ICE_INSET_TUN_IPV4_TTL;
+ input_set |= ICE_INSET_IPV4_TTL;
if (ipv4_mask->hdr.next_proto_id == UINT8_MAX)
- input_set |= ICE_INSET_TUN_IPV4_PROTO;
+ input_set |= ICE_INSET_IPV4_PROTO;
+ if (ipv4_mask->hdr.type_of_service == UINT8_MAX)
+ input_set |= ICE_INSET_IPV4_TOS;
}
break;
case RTE_FLOW_ITEM_TYPE_IPV6:
@@ -302,33 +300,32 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
return 0;
}
- if (outer_ip) {
+ if (is_tunnel) {
if (!memcmp(ipv6_mask->hdr.src_addr,
ipv6_addr_mask,
RTE_DIM(ipv6_mask->hdr.src_addr)))
- input_set |= ICE_INSET_IPV6_SRC;
+ input_set |= ICE_INSET_TUN_IPV6_SRC;
if (!memcmp(ipv6_mask->hdr.dst_addr,
ipv6_addr_mask,
RTE_DIM(ipv6_mask->hdr.dst_addr)))
- input_set |= ICE_INSET_IPV6_DST;
+ input_set |= ICE_INSET_TUN_IPV6_DST;
if (ipv6_mask->hdr.proto == UINT8_MAX)
- input_set |= ICE_INSET_IPV6_PROTO;
+ input_set |= ICE_INSET_TUN_IPV6_PROTO;
if (ipv6_mask->hdr.hop_limits == UINT8_MAX)
- input_set |= ICE_INSET_IPV6_HOP_LIMIT;
- outer_ip = false;
+ input_set |= ICE_INSET_TUN_IPV6_TTL;
} else {
if (!memcmp(ipv6_mask->hdr.src_addr,
ipv6_addr_mask,
RTE_DIM(ipv6_mask->hdr.src_addr)))
- input_set |= ICE_INSET_TUN_IPV6_SRC;
+ input_set |= ICE_INSET_IPV6_SRC;
if (!memcmp(ipv6_mask->hdr.dst_addr,
ipv6_addr_mask,
RTE_DIM(ipv6_mask->hdr.dst_addr)))
- input_set |= ICE_INSET_TUN_IPV6_DST;
+ input_set |= ICE_INSET_IPV6_DST;
if (ipv6_mask->hdr.proto == UINT8_MAX)
- input_set |= ICE_INSET_TUN_IPV6_PROTO;
+ input_set |= ICE_INSET_IPV6_PROTO;
if (ipv6_mask->hdr.hop_limits == UINT8_MAX)
- input_set |= ICE_INSET_TUN_IPV6_TTL;
+ input_set |= ICE_INSET_IPV6_HOP_LIMIT;
}
break;
@@ -353,17 +350,16 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
return 0;
}
- if (outer_l4) {
+ if (is_tunnel) {
if (udp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_SRC_PORT;
+ input_set |= ICE_INSET_TUN_SRC_PORT;
if (udp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_DST_PORT;
- outer_l4 = false;
+ input_set |= ICE_INSET_TUN_DST_PORT;
} else {
if (udp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_SRC_PORT;
+ input_set |= ICE_INSET_SRC_PORT;
if (udp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_DST_PORT;
+ input_set |= ICE_INSET_DST_PORT;
}
break;
@@ -393,17 +389,16 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
return 0;
}
- if (outer_l4) {
+ if (is_tunnel) {
if (tcp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_SRC_PORT;
+ input_set |= ICE_INSET_TUN_SRC_PORT;
if (tcp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_DST_PORT;
- outer_l4 = false;
+ input_set |= ICE_INSET_TUN_DST_PORT;
} else {
if (tcp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_SRC_PORT;
+ input_set |= ICE_INSET_SRC_PORT;
if (tcp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_DST_PORT;
+ input_set |= ICE_INSET_DST_PORT;
}
break;
@@ -427,17 +422,16 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
return 0;
}
- if (outer_l4) {
+ if (is_tunnel) {
if (sctp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_SRC_PORT;
+ input_set |= ICE_INSET_TUN_SRC_PORT;
if (sctp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_DST_PORT;
- outer_l4 = false;
+ input_set |= ICE_INSET_TUN_DST_PORT;
} else {
if (sctp_mask->hdr.src_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_SRC_PORT;
+ input_set |= ICE_INSET_SRC_PORT;
if (sctp_mask->hdr.dst_port == UINT16_MAX)
- input_set |= ICE_INSET_TUN_DST_PORT;
+ input_set |= ICE_INSET_DST_PORT;
}
break;
@@ -486,6 +480,7 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
"Invalid VXLAN item");
return 0;
}
+ is_tunnel = 1;
break;
case RTE_FLOW_ITEM_TYPE_NVGRE:
@@ -503,6 +498,7 @@ static uint64_t ice_get_flow_field(const struct rte_flow_item pattern[],
"Invalid NVGRE item");
return 0;
}
+ is_tunnel = 1;
break;
default:
--
2.9.5
^ permalink raw reply [flat|nested] 2+ messages in thread* [dpdk-stable] [PATCH v2] net/ice: fix flow validation fail issue
[not found] <20190715093807.128101-1-qiming.yang@intel.com>
2019-07-15 9:38 ` [dpdk-stable] [DPDK] net/ice: fix tunnel rule not recognize Qiming Yang
@ 2019-07-15 9:38 ` Qiming Yang
1 sibling, 0 replies; 2+ messages in thread
From: Qiming Yang @ 2019-07-15 9:38 UTC (permalink / raw)
To: qabuild; +Cc: Qiming Yang, stable
The return value of function ice_flow_valid_attr should
be non-zero. This patch fixed this issue.
Fixes: d76116a4678f ("net/ice: add generic flow API")
Cc: stable@dpdk.org
Signed-off-by: Qiming Yang <qiming.yang@intel.com>
---
drivers/net/ice/ice_generic_flow.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ice/ice_generic_flow.c b/drivers/net/ice/ice_generic_flow.c
index d5ff278..e6a2c4b 100644
--- a/drivers/net/ice/ice_generic_flow.c
+++ b/drivers/net/ice/ice_generic_flow.c
@@ -549,7 +549,7 @@ static int ice_flow_valid_action(struct rte_eth_dev *dev,
rte_flow_error_set(error, EINVAL,
RTE_FLOW_ERROR_TYPE_ACTION,
actions, "Invalid queue ID for"
- " ethertype_filter.");
+ " switch filter.");
return -rte_errno;
}
break;
@@ -596,7 +596,7 @@ ice_flow_validate(struct rte_eth_dev *dev,
}
ret = ice_flow_valid_attr(attr, error);
- if (!ret)
+ if (ret)
return ret;
inset = ice_flow_valid_pattern(pattern, error);
--
2.7.4
^ permalink raw reply [flat|nested] 2+ messages in thread