ethdev: fix last item detection on RSS flow expand
Checks
Commit Message
There is a rte_flow API which expands a RSS flow pattern to multipile
patterns according to the RSS hash types in the RSS action
configuration.
Aa part of the expansion, detection of the last item of the flow uses
the "next proto" field of the last configured item in the pattern list.
Wrongly, the mask of this field was not considered in order to validate
the field.
Ignore "next proto" fields when their corresponded masks invalidate them.
Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
Cc: jackmin@mellanox.com
Signed-off-by: Matan Azrad <matan@mellanox.com>
---
lib/librte_ethdev/rte_flow.c | 32 ++++++++++++++++++++++++++++++++
1 file changed, 32 insertions(+)
Comments
On Mon, 19-11-11, 10:42, Matan Azrad wrote:
> There is a rte_flow API which expands a RSS flow pattern to multipile
> patterns according to the RSS hash types in the RSS action
> configuration.
>
> Aa part of the expansion, detection of the last item of the flow uses
> the "next proto" field of the last configured item in the pattern list.
> Wrongly, the mask of this field was not considered in order to validate
> the field.
>
> Ignore "next proto" fields when their corresponded masks invalidate them.
>
> Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
> Cc: jackmin@mellanox.com
>
> Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Xiaoyu Min <jackmin@mellanox.com>
On 11/11/2019 10:42 AM, Matan Azrad wrote:
> There is a rte_flow API which expands a RSS flow pattern to multipile
> patterns according to the RSS hash types in the RSS action
> configuration.
>
> Aa part of the expansion, detection of the last item of the flow uses
> the "next proto" field of the last configured item in the pattern list.
> Wrongly, the mask of this field was not considered in order to validate
> the field.
>
> Ignore "next proto" fields when their corresponded masks invalidate them.
>
> Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
> Cc: jackmin@mellanox.com
>
> Signed-off-by: Matan Azrad <matan@mellanox.com>
Hi Ori,
Can you check this one please?
> -----Original Message-----
> From: dev <dev-bounces@dpdk.org> On Behalf Of Matan Azrad
> Sent: Monday, November 11, 2019 12:42 PM
> To: dev@dpdk.org
> Cc: Ori Kam <orika@mellanox.com>; Jack Min <jackmin@mellanox.com>
> Subject: [dpdk-dev] [PATCH] ethdev: fix last item detection on RSS flow
> expand
>
> There is a rte_flow API which expands a RSS flow pattern to multipile
> patterns according to the RSS hash types in the RSS action
> configuration.
>
> Aa part of the expansion, detection of the last item of the flow uses
> the "next proto" field of the last configured item in the pattern list.
> Wrongly, the mask of this field was not considered in order to validate
> the field.
>
> Ignore "next proto" fields when their corresponded masks invalidate them.
>
> Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
> Cc: jackmin@mellanox.com
>
> Signed-off-by: Matan Azrad <matan@mellanox.com>
> ---
Acked-by: Ori Kam <orika@mellanox.com>
Thanks,
Ori
Checked and acked.
Best,
Ori
> -----Original Message-----
> From: Ferruh Yigit <ferruh.yigit@intel.com>
> Sent: Monday, November 11, 2019 7:14 PM
> To: Matan Azrad <matan@mellanox.com>; dev@dpdk.org
> Cc: Ori Kam <orika@mellanox.com>; Jack Min <jackmin@mellanox.com>
> Subject: Re: [dpdk-dev] [PATCH] ethdev: fix last item detection on RSS flow
> expand
>
> On 11/11/2019 10:42 AM, Matan Azrad wrote:
> > There is a rte_flow API which expands a RSS flow pattern to multipile
> > patterns according to the RSS hash types in the RSS action
> > configuration.
> >
> > Aa part of the expansion, detection of the last item of the flow uses
> > the "next proto" field of the last configured item in the pattern list.
> > Wrongly, the mask of this field was not considered in order to validate
> > the field.
> >
> > Ignore "next proto" fields when their corresponded masks invalidate them.
> >
> > Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
> > Cc: jackmin@mellanox.com
> >
> > Signed-off-by: Matan Azrad <matan@mellanox.com>
>
> Hi Ori,
>
> Can you check this one please?
On 11/11/2019 5:41 PM, Ori Kam wrote:
>
>
>> -----Original Message-----
>> From: dev <dev-bounces@dpdk.org> On Behalf Of Matan Azrad
>> Sent: Monday, November 11, 2019 12:42 PM
>> To: dev@dpdk.org
>> Cc: Ori Kam <orika@mellanox.com>; Jack Min <jackmin@mellanox.com>
>> Subject: [dpdk-dev] [PATCH] ethdev: fix last item detection on RSS flow
>> expand
>>
>> There is a rte_flow API which expands a RSS flow pattern to multipile
>> patterns according to the RSS hash types in the RSS action
>> configuration.
>>
>> Aa part of the expansion, detection of the last item of the flow uses
>> the "next proto" field of the last configured item in the pattern list.
>> Wrongly, the mask of this field was not considered in order to validate
>> the field.
>>
>> Ignore "next proto" fields when their corresponded masks invalidate them.
>>
>> Fixes: ec84aa45f17b ("ethdev: fix expand RSS flows")
>> Cc: jackmin@mellanox.com
>>
>> Signed-off-by: Matan Azrad <matan@mellanox.com>
>> ---
>
> Acked-by: Ori Kam <orika@mellanox.com>
Applied to dpdk-next-net/master, thanks.
@@ -218,12 +218,21 @@ struct rte_flow_desc_data {
{
enum rte_flow_item_type ret = RTE_FLOW_ITEM_TYPE_VOID;
uint16_t ether_type = 0;
+ uint16_t ether_type_m;
uint8_t ip_next_proto = 0;
+ uint8_t ip_next_proto_m;
if (item == NULL || item->spec == NULL)
return ret;
switch (item->type) {
case RTE_FLOW_ITEM_TYPE_ETH:
+ if (item->mask)
+ ether_type_m = ((const struct rte_flow_item_eth *)
+ (item->mask))->type;
+ else
+ ether_type_m = rte_flow_item_eth_mask.type;
+ if (ether_type_m != RTE_BE16(0xFFFF))
+ break;
ether_type = ((const struct rte_flow_item_eth *)
(item->spec))->type;
if (rte_be_to_cpu_16(ether_type) == RTE_ETHER_TYPE_IPV4)
@@ -234,6 +243,13 @@ struct rte_flow_desc_data {
ret = RTE_FLOW_ITEM_TYPE_VLAN;
break;
case RTE_FLOW_ITEM_TYPE_VLAN:
+ if (item->mask)
+ ether_type_m = ((const struct rte_flow_item_vlan *)
+ (item->mask))->inner_type;
+ else
+ ether_type_m = rte_flow_item_vlan_mask.inner_type;
+ if (ether_type_m != RTE_BE16(0xFFFF))
+ break;
ether_type = ((const struct rte_flow_item_vlan *)
(item->spec))->inner_type;
if (rte_be_to_cpu_16(ether_type) == RTE_ETHER_TYPE_IPV4)
@@ -244,6 +260,14 @@ struct rte_flow_desc_data {
ret = RTE_FLOW_ITEM_TYPE_VLAN;
break;
case RTE_FLOW_ITEM_TYPE_IPV4:
+ if (item->mask)
+ ip_next_proto_m = ((const struct rte_flow_item_ipv4 *)
+ (item->mask))->hdr.next_proto_id;
+ else
+ ip_next_proto_m =
+ rte_flow_item_ipv4_mask.hdr.next_proto_id;
+ if (ip_next_proto_m != 0xFF)
+ break;
ip_next_proto = ((const struct rte_flow_item_ipv4 *)
(item->spec))->hdr.next_proto_id;
if (ip_next_proto == IPPROTO_UDP)
@@ -256,6 +280,14 @@ struct rte_flow_desc_data {
ret = RTE_FLOW_ITEM_TYPE_IPV6;
break;
case RTE_FLOW_ITEM_TYPE_IPV6:
+ if (item->mask)
+ ip_next_proto_m = ((const struct rte_flow_item_ipv6 *)
+ (item->mask))->hdr.proto;
+ else
+ ip_next_proto_m =
+ rte_flow_item_ipv6_mask.hdr.proto;
+ if (ip_next_proto_m != 0xFF)
+ break;
ip_next_proto = ((const struct rte_flow_item_ipv6 *)
(item->spec))->hdr.proto;
if (ip_next_proto == IPPROTO_UDP)