Skip to contentSkip to conversation

ethdev: update qfi definition

PATCH Applied Raslan Darawsheh 6 participants 37 messages Mar 23, 2021 – Jun 8, 2021

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ethdev: update qfi definition

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 3 ++-
 lib/librte_ethdev/rte_flow.h                | 4 ++--
 2 files changed, 4 insertions(+), 3 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
 - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
   - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
 - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +2 -2
index 669e677e91..79106e0246 100644
@@ -1392,14 +1392,14 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
  */
 struct rte_flow_item_gtp_psc {
 	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
 };
 
 /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
 #ifndef __cplusplus
 static const struct rte_flow_item_gtp_psc
 rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.qfi = 0xff,
 };
 #endif
 
Signature · 1 line
2.17.1
#1 lore raw
Quoted 4 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

Can you please put a reference for above information?

Quoted 36 lines

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst | 3 ++-
  lib/librte_ethdev/rte_flow.h                | 4 ++--
  2 files changed, 4 insertions(+), 3 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +0 -0
index 669e677e91..79106e0246 100644

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1396

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +1 -1
@@ -1392,14 +1392,14 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
   */
  struct rte_flow_item_gtp_psc {
  	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
  };

By design requirement, rte_flow_item_* should start with the relevant protocol header.

There is already a deprecation notice for using protocol header directly in the rte_flow_item* [1] and Adrew/Ivan already fixed a few of them [2].

Your patch is not directly related on this, but since you are touching to the flow_item, would you mind create a protocol struct for it first, and use it in the "struct rte_flow_item_gtp_psc"? So the protocol related update can be done in the protocol header, instead of rte_flow struct.

[1]
https://git.dpdk.org/dpdk/tree/doc/guides/rel_notes/deprecation.rst?h=v21.02#n99

[2]
https://git.dpdk.org/next/dpdk-next-net/commit/?id=4a061a7bd70bfa023de23e8e654e

#2 lore raw

Hi

Quoted 60 lines

-----Original Message-----
From: dev <[email protected]> On Behalf Of Ferruh Yigit
Subject: Re: [dpdk-dev] [PATCH] ethdev: update qfi definition

Quoted 4 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

Can you please put a reference for above information?

Quoted 11 lines

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst | 3 ++-
  lib/librte_ethdev/rte_flow.h                | 4 ++--
  2 files changed, 4 insertions(+), 3 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and their

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their

attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 14 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +0 -0
index 669e677e91..79106e0246 100644
@@ -1392,14 +1392,14 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1396

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +1 -1
   */
  struct rte_flow_item_gtp_psc {
  	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
  };

By design requirement, rte_flow_item_* should start with the relevant protocol header.

Quoted 9 lines

There is already a deprecation notice for using protocol header directly in the rte_flow_item* [1] and Adrew/Ivan already fixed a few of them [2].

Your patch is not directly related on this, but since you are touching to the flow_item, would you mind create a protocol struct for it first, and use it in the "struct rte_flow_item_gtp_psc"?
So the protocol related update can be done in the protocol header, instead of rte_flow struct.

+1

Quoted 20 lines

[1]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.dpdk.
org%2Fdpdk%2Ftree%2Fdoc%2Fguides%2Frel_notes%2Fdeprecation.rst%3Fh%3 Dv21.02%23n99&amp;data=04%7C01%7Corika%40nvidia.com%7C6391a4c0640 f4592b70b08d8f058e322%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0 %7C637523611870497932%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAw MDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdat a=vU%2F5oO47zb9erTnZL1pl9j0nHCKzea3NJgOeo1FTW0Q%3D&amp;reserved= 0

[2]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.dpdk.
org%2Fnext%2Fdpdk-next- net%2Fcommit%2F%3Fid%3D4a061a7bd70bfa023de23e8e654e&amp;data=04% 7C01%7Corika%40nvidia.com%7C6391a4c0640f4592b70b08d8f058e322%7C43 083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C637523611870497932%7CU nknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik 1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=WurlQ5VSLqFGVthfRj363xZsC No3xJuvxNQCFVcxdkk%3D&amp;reserved=0

#3 lore raw

Hi,

Quoted 12 lines

-----Original Message-----
From: Ori Kam <[email protected]>
Sent: Monday, March 29, 2021 11:53 AM
To: Ferruh Yigit <[email protected]>; Raslan Darawsheh <[email protected]>; [email protected]; Andrew Rybchenko <[email protected]>; Ivan Malov <[email protected]>
Cc: Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; [email protected]; NBU- Contact-Thomas Monjalon <[email protected]>; Olivier Matz <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: update qfi definition

Quoted 14 lines

Hi

Quoted 12 lines

-----Original Message-----
From: dev <[email protected]> On Behalf Of Ferruh Yigit
Subject: Re: [dpdk-dev] [PATCH] ethdev: update qfi definition

Quoted 4 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

Can you please put a reference for above information?

Quoted 1 line

Sure, will send in V2,

Quoted 54 lines
Quoted 16 lines
Quoted 16 lines

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst | 3 ++-
  lib/librte_ethdev/rte_flow.h                | 4 ++--
  2 files changed, 4 insertions(+), 3 deletions(-)
Quoted 6 lines

Review

Quoted 1 line
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their

Quoted 16 lines

attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 14 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +0 -0
index 669e677e91..79106e0246 100644
@@ -1392,14 +1392,14 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 8 lines

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1396

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +1 -1
   */
  struct rte_flow_item_gtp_psc {
  	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
  };

By design requirement, rte_flow_item_* should start with the relevant

protocol

Quoted 3 lines

header.

There is already a deprecation notice for using protocol header directly in

the

Quoted 2 lines

rte_flow_item* [1] and Adrew/Ivan already fixed a few of them [2].

Quoted 8 lines
Quoted 4 lines

Your patch is not directly related on this, but since you are touching to the flow_item, would you mind create a protocol struct for it first, and use it in the "struct rte_flow_item_gtp_psc"?
So the protocol related update can be done in the protocol header, instead

of

Quoted 2 lines

rte_flow struct.

+1

Sure, I can create the new protocol and use it , and will send in V2. But, wouldn't it cause any ABI breakage ?

Quoted 44 lines
Quoted 4 lines

[1]

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.d
pdk.

Quoted 1 line

org%2Fdpdk%2Ftree%2Fdoc%2Fguides%2Frel_notes%2Fdeprecation.rst%3F h%3

Quoted 1 line

Dv21.02%23n99&amp;data=04%7C01%7Corika%40nvidia.com%7C6391a4c064 0

Quoted 1 line

f4592b70b08d8f058e322%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0

Quoted 1 line

%7C637523611870497932%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjA w

Quoted 1 line

MDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sda t

Quoted 1 line

a=vU%2F5oO47zb9erTnZL1pl9j0nHCKzea3NJgOeo1FTW0Q%3D&amp;reserve d=

Quoted 4 lines

0

[2]

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.d
pdk.

Quoted 2 lines

org%2Fnext%2Fdpdk-next-

net%2Fcommit%2F%3Fid%3D4a061a7bd70bfa023de23e8e654e&amp;data=0 4%

Quoted 1 line

7C01%7Corika%40nvidia.com%7C6391a4c0640f4592b70b08d8f058e322%7C43

Quoted 1 line

083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C637523611870497932%7CU

Quoted 1 line

nknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI 6Ik

Quoted 1 line

1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=WurlQ5VSLqFGVthfRj363xZs C

Quoted 1 line

No3xJuvxNQCFVcxdkk%3D&amp;reserved=0

#4 lore raw
Quoted 88 lines · Raslan Darawsheh

Hi,

Quoted 12 lines

-----Original Message-----
From: Ori Kam <[email protected]>
Sent: Monday, March 29, 2021 11:53 AM
To: Ferruh Yigit <[email protected]>; Raslan Darawsheh <[email protected]>; [email protected]; Andrew Rybchenko <[email protected]>; Ivan Malov <[email protected]>
Cc: Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; [email protected]; NBU- Contact-Thomas Monjalon <[email protected]>; Olivier Matz <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: update qfi definition

Quoted 14 lines

Hi

Quoted 12 lines

-----Original Message-----
From: dev <[email protected]> On Behalf Of Ferruh Yigit
Subject: Re: [dpdk-dev] [PATCH] ethdev: update qfi definition

Quoted 4 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

Can you please put a reference for above information?

Quoted 1 line

Sure, will send in V2,

Quoted 58 lines
Quoted 16 lines
Quoted 16 lines

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
   doc/guides/testpmd_app_ug/testpmd_funcs.rst | 3 ++-
   lib/librte_ethdev/rte_flow.h                | 4 ++--
   2 files changed, 4 insertions(+), 3 deletions(-)
Quoted 6 lines

Review

Quoted 1 line
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their

Quoted 16 lines

attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 14 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
   - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
     - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
   - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +0 -0
index 669e677e91..79106e0246 100644
@@ -1392,14 +1392,14 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 8 lines

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1396

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +1 -1
    */
   struct rte_flow_item_gtp_psc {
   	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
   };

By design requirement, rte_flow_item_* should start with the relevant

protocol

Quoted 3 lines

header.

There is already a deprecation notice for using protocol header directly in

the

Quoted 6 lines

rte_flow_item* [1] and Adrew/Ivan already fixed a few of them [2].

Your patch is not directly related on this, but since you are touching to the flow_item, would you mind create a protocol struct for it first, and use it in the "struct rte_flow_item_gtp_psc"?
So the protocol related update can be done in the protocol header, instead

Quoted 5 lines
Quoted 4 lines

of

Quoted 2 lines

rte_flow struct.

+1

Sure, I can create the new protocol and use it , and will send in V2. But, wouldn't it cause any ABI breakage ?

For the protocol header, it won't be problem since it is a new struct.

For the flow item struct, it may be, that is why need to use a union for now, please check what Adrew/Ivan did.

Quoted 45 lines
Quoted 44 lines
Quoted 4 lines

[1]

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.d
pdk.

Quoted 1 line

org%2Fdpdk%2Ftree%2Fdoc%2Fguides%2Frel_notes%2Fdeprecation.rst%3F h%3

Quoted 1 line

Dv21.02%23n99&amp;data=04%7C01%7Corika%40nvidia.com%7C6391a4c064 0

Quoted 1 line

f4592b70b08d8f058e322%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0

Quoted 1 line

%7C637523611870497932%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjA w

Quoted 1 line

MDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sda t

Quoted 1 line

a=vU%2F5oO47zb9erTnZL1pl9j0nHCKzea3NJgOeo1FTW0Q%3D&amp;reserve d=

Quoted 4 lines

0

[2]

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.d
pdk.

Quoted 2 lines

org%2Fnext%2Fdpdk-next-

net%2Fcommit%2F%3Fid%3D4a061a7bd70bfa023de23e8e654e&amp;data=0 4%

Quoted 1 line

7C01%7Corika%40nvidia.com%7C6391a4c0640f4592b70b08d8f058e322%7C43

Quoted 1 line

083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C637523611870497932%7CU

Quoted 1 line

nknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI 6Ik

Quoted 1 line

1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=WurlQ5VSLqFGVthfRj363xZs C

Quoted 1 line

No3xJuvxNQCFVcxdkk%3D&amp;reserved=0

#5 lore raw

v2 cover fix gtp psc qfi support

In reply to Raslan Darawsheh

This is fixing the gtp psc support to match the RFC 38415-g30

v2: introduce new header definition for gtp psc
update commit msg for rte flow item change.

Raslan Darawsheh (2):
ethdev: add new ext hdr for gtp psc
ethdev: update qfi definition

 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 +-
 lib/librte_ethdev/rte_flow.h                | 18 +++++++++--
 lib/librte_net/rte_gtp.h                    | 34 +++++++++++++++++++++
 3 files changed, 51 insertions(+), 4 deletions(-)
Signature · 1 line
2.29.0
#6 lore raw

PATCH v2 1/2 ethdev: add new ext hdr for gtp psc

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +34 -0
index 6a6f9b238d..8af48ea1ec 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extention for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence */
+			uint8_t ul_delay_ind:1; /**< ul delay result presence */
+			uint8_t snp_ul1:1; /**< Sequence number presence ul */
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
Signature · 1 line
2.29.0
#7 lore raw

v3 cover fix gtp psc qfi support

This is fixing the gtp psc support to match the RFC 38415-g30

v2: introduce new header definition for gtp psc
update commit msg for rte flow item change.

v3: fixed typo in comment
Cc relevant people.

Raslan Darawsheh (2):
ethdev: add new ext hdr for gtp psc
ethdev: update qfi definition

 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 +-
 lib/librte_ethdev/rte_flow.h                | 18 +++++++++--
 lib/librte_net/rte_gtp.h                    | 34 +++++++++++++++++++++
 3 files changed, 51 insertions(+), 4 deletions(-)
Signature · 1 line
2.29.0
#8 lore raw

PATCH v3 1/2 ethdev: add new ext hdr for gtp psc

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +34 -0
index 6a6f9b238d..bfaae26535 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence */
+			uint8_t ul_delay_ind:1; /**< ul delay result presence */
+			uint8_t snp_ul1:1; /**< Sequence number presence ul */
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
Signature · 1 line
2.29.0
#9 lore raw

Re: [dpdk-dev] [PATCH v3 1/2] ethdev: add new ext hdr for gtp psc

Quoted 13 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
  lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
  1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +0 -0
index 6a6f9b238d..bfaae26535 100644

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 69

Quoted 9 lines
lib/librte_net/rte_gtp.h +6 -0
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
  	uint8_t next_ext;     /**< Next Extension Header Type. */
  }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc {

general consensus in other protocols seems having '_hdr' suffix in the struct name, I can see this is extension but do you think does it make to add suffix?

Quoted 32 lines
lib/librte_net/rte_gtp.h +28 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence */
+			uint8_t ul_delay_ind:1; /**< ul delay result presence */
+			uint8_t snp_ul1:1; /**< Sequence number presence ul */
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;
+
  /** GTP header length */
  #define RTE_ETHER_GTP_HLEN \
  	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
#10 lore raw

Re: [dpdk-dev] [PATCH v3 1/2] ethdev: add new ext hdr for gtp psc

Quoted 27 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, April 1, 2021 7:51 PM
To: Raslan Darawsheh <[email protected]>; [email protected]
Cc: Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v3 1/2] ethdev: add new ext hdr for gtp psc

Quoted 16 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
  lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
  1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +0 -0
index 6a6f9b238d..bfaae26535 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
  	uint8_t next_ext;     /**< Next Extension Header Type. */
  }  __rte_packed;
 
Quoted 9 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 69

Quoted 6 lines
lib/librte_net/rte_gtp.h +6 -0
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc {

general consensus in other protocols seems having '_hdr' suffix in the struct name, I can see this is extension but do you think does it make to add suffix?

Yes sure, will send in a new version.

Quoted 26 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 11 lines
lib/librte_net/rte_gtp.h +10 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 80

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t ul_delay_ind:1; /**< ul delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 81

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t snp_ul1:1; /**< Sequence number presence

ul */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result

presence */

Quoted 11 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 100

Quoted 11 lines
lib/librte_net/rte_gtp.h +7 -0
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;
+
  /** GTP header length */
  #define RTE_ETHER_GTP_HLEN \
  	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))

Kindest regards
Raslan Darawsheh

#11 lore raw

PATCH v3 2/2 ethdev: update qfi definition

In reply to Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
 lib/librte_ethdev/rte_flow.h                | 18 +++++++++++++++---
 2 files changed, 17 insertions(+), 4 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
 - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
   - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
 - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +15 -3
index 6cc57136ac..1eb9711707 100644
@@ -20,6 +20,7 @@
 #include <rte_arp.h>
 #include <rte_common.h>
 #include <rte_ether.h>
+#include <rte_gtp.h>
 #include <rte_icmp.h>
 #include <rte_ip.h>
 #include <rte_sctp.h>
@@ -1421,16 +1422,27 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
  *
  * Matches a GTP PDU extension header with type 0x85.
  */
+RTE_STD_C11
 struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc gtp_psc;
+	};
 };
 
 /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
 #ifndef __cplusplus
 static const struct rte_flow_item_gtp_psc
 rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.qfi = 0xff,
 };
 #endif
 
Signature · 1 line
2.29.0
#12 lore raw
Quoted 56 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
  lib/librte_ethdev/rte_flow.h                | 18 +++++++++++++++---
  2 files changed, 17 insertions(+), 4 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +3 -2
index 6cc57136ac..1eb9711707 100644
@@ -20,6 +20,7 @@
  #include <rte_arp.h>
  #include <rte_common.h>
  #include <rte_ether.h>
+#include <rte_gtp.h>
  #include <rte_icmp.h>
  #include <rte_ip.h>
  #include <rte_sctp.h>
@@ -1421,16 +1422,27 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
   *
   * Matches a GTP PDU extension header with type 0x85.
   */
+RTE_STD_C11
  struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1437

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +10 -0
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc gtp_psc;

Again for consistency, what do you think to rename the variable as 'hdr'?

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1445

Quoted 9 lines
lib/librte_ethdev/rte_flow.h +2 -1
+	};
  };
 
  /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
  #ifndef __cplusplus
  static const struct rte_flow_item_gtp_psc
  rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.qfi = 0xff,

Since the protocol header is the preferred way, (individual fields may be deprecated in the future), can you please switch to new field, like:
.gtp_psc.qfi = 0xff,

Quoted 4 lines
lib/librte_ethdev/rte_flow.h +0 -0
  };
  #endif
 
#13 lore raw

Hi Ferruh,

Quoted 74 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, April 1, 2021 7:54 PM
To: Raslan Darawsheh <[email protected]>; [email protected]
Cc: Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]
Subject: Re: [PATCH v3 2/2] ethdev: update qfi definition

Quoted 15 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
  lib/librte_ethdev/rte_flow.h                | 18 +++++++++++++++---
  2 files changed, 17 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 22 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +1 -0
index 6cc57136ac..1eb9711707 100644
@@ -20,6 +20,7 @@
  #include <rte_arp.h>
  #include <rte_common.h>
  #include <rte_ether.h>
+#include <rte_gtp.h>
  #include <rte_icmp.h>
  #include <rte_ip.h>
  #include <rte_sctp.h>
@@ -1421,16 +1422,27 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 12 lines
lib/librte_ethdev/rte_flow.h +6 -2
   *
   * Matches a GTP PDU extension header with type 0x85.
   */
+RTE_STD_C11
  struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
Quoted 8 lines

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1437

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +6 -0
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc gtp_psc;

Again for consistency, what do you think to rename the variable as 'hdr'?

Will do,

Quoted 4 lines
Quoted 4 lines
lib/librte_ethdev/rte_flow.h +1 -0
+	};
  };
 
Quoted 9 lines

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1445

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +1 -1
  /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
  #ifndef __cplusplus
  static const struct rte_flow_item_gtp_psc
  rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.qfi = 0xff,

Since the protocol header is the preferred way, (individual fields may be deprecated in the future), can you please switch to new field, like:

Yes, will do it accordingly

Quoted 6 lines

.gtp_psc.qfi = 0xff,

Quoted 4 lines
lib/librte_ethdev/rte_flow.h +0 -0
  };
  #endif
 
#14 lore raw

v4 cover fix gtp psc qfi support

In reply to Raslan Darawsheh

This is fixing the gtp psc support to match the RFC 38415-g30

v2: introduce new header definition for gtp psc
update commit msg for rte flow item change.

v3: fixed typo in comment
Cc relevant people.

v4: update hdr definition to have hdr suffix.
update variable name to be hdr in the gtp_psc item. update default max to use the new added hdr.

Raslan Darawsheh (2):
ethdev: add new ext hdr for gtp psc
ethdev: update qfi definition

 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 +-
 lib/librte_ethdev/rte_flow.h                | 20 ++++++++++--
 lib/librte_net/rte_gtp.h                    | 34 +++++++++++++++++++++
 3 files changed, 53 insertions(+), 4 deletions(-)
Signature · 1 line
2.29.0
#15 lore raw

PATCH v4 1/2 ethdev: add new ext hdr for gtp psc

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +34 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence */
+			uint8_t ul_delay_ind:1; /**< ul delay result presence */
+			uint8_t snp_ul1:1; /**< Sequence number presence ul */
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
Signature · 1 line
2.29.0
#16 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Hi Raslan,

Quoted 2 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

Do you have a link to this RFC?

Quoted 36 lines
---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +23 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence */
+			uint8_t ul_delay_ind:1; /**< ul delay result presence */
+			uint8_t snp_ul1:1; /**< Sequence number presence ul */
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 96

Quoted 10 lines
lib/librte_net/rte_gtp.h +10 -0
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;

With this header, sizeof(rte_gtp_psc_hdr) = 5, is it really expected?

It would help to see the specification to have a better idea of how to split, but a possible solution is to do something like this:

struct rte_gtp_psc_generic_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
};
struct rte_gtp_psc_<name1>_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t uint8_t snp:1;
	uint8_t spare_dl1:2;
	...
};

...

struct rte_gtp_psc_hdr {
	union {
		struct rte_gtp_psc_generic_hdr generic;
		struct rte_gtp_psc_<name1>_hdr <name1>;
		struct rte_gtp_psc_<name2>_hdr <name2>;
	};
};

Also, you need to take care about endianness.

Regards,
Olivier

#17 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Hi Olivier,

Quoted 10 lines

-----Original Message-----
From: Olivier Matz <[email protected]>
Sent: Thursday, April 8, 2021 3:30 PM
To: Raslan Darawsheh <[email protected]>
Cc: [email protected]; [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 7 lines

Hi Raslan,

Quoted 2 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

Do you have a link to this RFC?

Yes sure,
https://www.3gpp.org/ftp/Specs/archive/38_series/38.415/38415-g30.zip

Quoted 45 lines
Quoted 20 lines
---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +6 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 79

Quoted 10 lines
lib/librte_net/rte_gtp.h +10 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 80

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t ul_delay_ind:1; /**< ul delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 81

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t snp_ul1:1; /**< Sequence number presence

ul */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result

presence */

Quoted 8 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 96

Quoted 6 lines
lib/librte_net/rte_gtp.h +6 -0
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;

With this header, sizeof(rte_gtp_psc_hdr) = 5, is it really expected?

The data[0] is variable length data, I guess I should send another version to mention that in the comment maybe. The header size according to the spec should be 4 octets aligned in general.

Quoted 2 lines

It would help to see the specification to have a better idea of how to

Sure, I've just posted the link above, please let me know of any suggestion that you have, and I'll be glad to do accordingly.

Quoted 28 lines

split, but a possible solution is to do something like this:

struct rte_gtp_psc_generic_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
};
struct rte_gtp_psc_<name1>_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t uint8_t snp:1;
	uint8_t spare_dl1:2;
	...
};

...

struct rte_gtp_psc_hdr {
	union {
		struct rte_gtp_psc_generic_hdr generic;
		struct rte_gtp_psc_<name1>_hdr <name1>;
		struct rte_gtp_psc_<name2>_hdr <name2>;
	};
};
Quoted 5 lines

Also, you need to take care about endianness.

Regards,
Olivier

Kindest regards
Raslan Darawsheh

#18 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 75 lines · Raslan Darawsheh

Hi Olivier,

Quoted 10 lines

-----Original Message-----
From: Olivier Matz <[email protected]>
Sent: Thursday, April 8, 2021 3:30 PM
To: Raslan Darawsheh <[email protected]>
Cc: [email protected]; [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 7 lines

Hi Raslan,

Quoted 2 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

Do you have a link to this RFC?

Yes sure,
https://www.3gpp.org/ftp/Specs/archive/38_series/38.415/38415-g30.zip

Quoted 45 lines
Quoted 20 lines
---
  lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
  1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +6 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
  	uint8_t next_ext;     /**< Next Extension Header Type. */
  }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 79

Quoted 10 lines
lib/librte_net/rte_gtp.h +10 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 80

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t ul_delay_ind:1; /**< ul delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 81

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t snp_ul1:1; /**< Sequence number presence

ul */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result

presence */

Quoted 8 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 96

Quoted 6 lines
lib/librte_net/rte_gtp.h +6 -0
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;

With this header, sizeof(rte_gtp_psc_hdr) = 5, is it really expected?

Quoted 2 lines

The data[0] is variable length data, I guess I should send another version to mention that in the comment maybe. The header size according to the spec should be 4 octets aligned in general.

The struct is 5 btyes, this is not related to data[0], please check the pahole output:
http://inbox.dpdk.org/dev/[email protected]/

Quoted 40 lines
Quoted 2 lines

It would help to see the specification to have a better idea of how to

Sure, I've just posted the link above, please let me know of any suggestion that you have, and I'll be glad to do accordingly.

Quoted 28 lines

split, but a possible solution is to do something like this:

struct rte_gtp_psc_generic_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
};
struct rte_gtp_psc_<name1>_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t uint8_t snp:1;
	uint8_t spare_dl1:2;
	...
};

...

struct rte_gtp_psc_hdr {
	union {
		struct rte_gtp_psc_generic_hdr generic;
		struct rte_gtp_psc_<name1>_hdr <name1>;
		struct rte_gtp_psc_<name2>_hdr <name2>;
	};
};
Quoted 5 lines

Also, you need to take care about endianness.

Regards,
Olivier

Kindest regards
Raslan Darawsheh

#19 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

In reply to Raslan Darawsheh
Quoted 75 lines · Raslan Darawsheh

Hi Olivier,

Quoted 10 lines

-----Original Message-----
From: Olivier Matz <[email protected]>
Sent: Thursday, April 8, 2021 3:30 PM
To: Raslan Darawsheh <[email protected]>
Cc: [email protected]; [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 7 lines

Hi Raslan,

Quoted 2 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

Do you have a link to this RFC?

Yes sure,
https://www.3gpp.org/ftp/Specs/archive/38_series/38.415/38415-g30.zip

Quoted 45 lines
Quoted 20 lines
---
 lib/librte_net/rte_gtp.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +6 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 79

Quoted 10 lines
lib/librte_net/rte_gtp.h +10 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 80

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t ul_delay_ind:1; /**< ul delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 81

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t snp_ul1:1; /**< Sequence number presence

ul */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result

presence */

Quoted 8 lines

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 96

Quoted 6 lines
lib/librte_net/rte_gtp.h +6 -0
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;

With this header, sizeof(rte_gtp_psc_hdr) = 5, is it really expected?

Quoted 2 lines

The data[0] is variable length data, I guess I should send another version to mention that in the comment maybe. The header size according to the spec should be 4 octets aligned in general.

What I wanted to highlight is that using union of structs containing bitfields does not work as you expect: each union is at least 1 byte. This results in a structure that does not match the expected header.

Quoted 31 lines
Quoted 2 lines

It would help to see the specification to have a better idea of how to

Sure, I've just posted the link above, please let me know of any suggestion that you have, and I'll be glad to do accordingly.

Quoted 27 lines

split, but a possible solution is to do something like this:

struct rte_gtp_psc_generic_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
};
struct rte_gtp_psc_<name1>_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t uint8_t snp:1;
	uint8_t spare_dl1:2;
	...
};

...

struct rte_gtp_psc_hdr {
	union {
		struct rte_gtp_psc_generic_hdr generic;
		struct rte_gtp_psc_<name1>_hdr <name1>;
		struct rte_gtp_psc_<name2>_hdr <name2>;
	};
};

From what I see in the documation, I think this approach should work. From afar, I suggest:

struct rte_gtp_psc_generic_hdr {

#if big endian
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
#else
	uint8_t pad:3;
	uint8_t qmp:1;
	uint8_t type:4

#endif
};

struct rte_gtp_psc_type0_hdr {

#if big endian
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t snp:1;
	uint8_t spare:2;

	uint8_t ppp:1;
	...
#else
	uint8_t pad:3;
	uint8_t qmp:1;
	uint8_t type:4
	uint8_t spare:2;
	uint8_t snp:1;

	...

#endif
uint8_t data[0]; /* for variable fields */
};

struct rte_gtp_psc_type1_hdr {
... same for fixed fields of type1

uint8_t data[0]; /* for variable fields */
};

I don't see in the spec where is the reference to ext_hdr_len.

Regards,
Olivier

#20 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Hi and sorry for late response,

Quoted 95 lines

-----Original Message-----
From: Olivier Matz <[email protected]>
Sent: Thursday, April 8, 2021 5:11 PM
To: Raslan Darawsheh <[email protected]>
Cc: [email protected]; [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 12 lines · Raslan Darawsheh

Hi Olivier,

Quoted 10 lines

-----Original Message-----
From: Olivier Matz <[email protected]>
Sent: Thursday, April 8, 2021 3:30 PM
To: Raslan Darawsheh <[email protected]>
Cc: [email protected]; [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 9 lines
Quoted 7 lines

Hi Raslan,

Quoted 2 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

Do you have a link to this RFC?

Yes sure,

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fww
w.3gpp.org%2Fftp%2FSpecs%2Farchive%2F38_series%2F38.415%2F38415- g30.zip&amp;data=04%7C01%7Crasland%40nvidia.com%7C5279dde172024bc d1f6b08d8fa981668%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C6 37534878435191995%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMD AiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata= LuD7%2FvJgqJbZ78ncaM7UrpiLNPUt7zbPPfSMemyb2Y8%3D&amp;reserved =0

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
---
 lib/librte_net/rte_gtp.h | 34

++++++++++++++++++++++++++++++++++

Quoted 49 lines
Quoted 49 lines
Quoted 16 lines
 1 file changed, 34 insertions(+)
lib/librte_net/rte_gtp.h +6 -0
index 6a6f9b238d..088b0b5a53 100644
@@ -61,6 +61,40 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 79

Quoted 10 lines
lib/librte_net/rte_gtp.h +10 -0
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	union {
+		struct {
+			uint8_t snp:1; /**< Sequence number presence */
+			uint8_t spare_dl1:2; /**< spare down link bits */
+		};
+		struct {
+			uint8_t dl_delay_ind:1; /**< dl delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 80

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t ul_delay_ind:1; /**< ul delay result presence

*/

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 81

Quoted 1 line
lib/librte_net/rte_gtp.h +1 -0
+			uint8_t snp_ul1:1; /**< Sequence number presence

ul */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word {

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+		};
+	};
+	union {
+		struct {
+			uint8_t ppp:1; /**< Paging policy presence */
+			uint8_t rqi:1; /**< Reflective Qos Indicator */
+		};
+		struct {
+			uint8_t n_delay_ind:1; /**< N3/N9 delay result

presence */

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 96

Quoted 6 lines
lib/librte_net/rte_gtp.h +6 -0
+			uint8_t spare_ul2:1; /**< spare up link bits */
+		};
+	};
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t data[0]; /**< data feilds */
+} __rte_packed;

With this header, sizeof(rte_gtp_psc_hdr) = 5, is it really expected?

Quoted 7 lines
Quoted 1 line

The data[0] is variable length data, I guess I should send another version to

mention that in the comment maybe.

Quoted 1 line

The header size according to the spec should be 4 octets aligned in general.

What I wanted to highlight is that using union of structs containing bitfields does not work as you expect: each union is at least 1 byte. This results in a structure that does not match the expected header.

I see thanks for explaining.

Quoted 70 lines
Quoted 4 lines
Quoted 3 lines

It would help to see the specification to have a better idea of how to

Sure, I've just posted the link above, please let me know of any suggestion

that you have, and I'll be glad to do accordingly.

Quoted 21 lines
Quoted 21 lines

split, but a possible solution is to do something like this:

struct rte_gtp_psc_generic_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
};
struct rte_gtp_psc_<name1>_hdr {
	uint8_t ext_hdr_len;
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t uint8_t snp:1;
	uint8_t spare_dl1:2;
	...
};

...

Quoted 7 lines
Quoted 7 lines
struct rte_gtp_psc_hdr {
	union {
		struct rte_gtp_psc_generic_hdr generic;
		struct rte_gtp_psc_<name1>_hdr <name1>;
		struct rte_gtp_psc_<name2>_hdr <name2>;
	};
};

From what I see in the documation, I think this approach should work. From afar, I suggest:

struct rte_gtp_psc_generic_hdr {

#if big endian
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t pad:3;
#else
	uint8_t pad:3;
	uint8_t qmp:1;
	uint8_t type:4

#endif
};

struct rte_gtp_psc_type0_hdr {

#if big endian
	uint8_t type:4
	uint8_t qmp:1;
	uint8_t snp:1;
	uint8_t spare:2;

	uint8_t ppp:1;
	...
#else
	uint8_t pad:3;
	uint8_t qmp:1;
	uint8_t type:4
	uint8_t spare:2;
	uint8_t snp:1;

	...

#endif
uint8_t data[0]; /* for variable fields */
};

Quoted 7 lines

struct rte_gtp_psc_type1_hdr {
... same for fixed fields of type1

uint8_t data[0]; /* for variable fields */
};

Sure, will consider and do it this way in the next version.

Quoted 2 lines

I don't see in the spec where is the reference to ext_hdr_len.

The ext_hdr_len is part of the definition for all ext_hdrs see this link for more details:
https://en.wikipedia.org/wiki/GPRS_Tunnelling_Protocol
So, it's not mentioned in the doc.

Quoted 2 lines

Regards,
Olivier

#21 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

In reply to Olivier Matz
Quoted 11 lines · Olivier Matz

Hi Raslan,

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 72

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */

would it be a lot ot ask to have the structure defined using standard C instead of using compiler specific extensions?

#22 lore raw

Re: [dpdk-dev] [PATCH v4 1/2] ethdev: add new ext hdr for gtp psc

Quoted 16 lines · Tyler Retzlaff
Quoted 11 lines · Olivier Matz

Hi Raslan,

lib/librte_net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 72

Quoted 9 lines
lib/librte_net/rte_gtp.h +9 -0
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */

would it be a lot ot ask to have the structure defined using standard C instead of using compiler specific extensions?

It is a valid request, but I'm afraid bit fields are used in many-many core DPDK libraries. So, change of the approach or at least direction for a new code should be discussed and approved by techboard.

#23 lore raw

PATCH v4 2/2 ethdev: update qfi definition

In reply to Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
 lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
 2 files changed, 19 insertions(+), 4 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
 - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
   - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
 - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +17 -3
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
 #include <rte_arp.h>
 #include <rte_common.h>
 #include <rte_ether.h>
+#include <rte_gtp.h>
 #include <rte_icmp.h>
 #include <rte_ip.h>
 #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
  *
  * Matches a GTP PDU extension header with type 0x85.
  */
+RTE_STD_C11
 struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
 };
 
 /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
 #ifndef __cplusplus
 static const struct rte_flow_item_gtp_psc
 rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
 };
 #endif
 
Signature · 1 line
2.29.0
#24 lore raw
Quoted 58 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
  lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
  2 files changed, 19 insertions(+), 4 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +5 -2
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
  #include <rte_arp.h>
  #include <rte_common.h>
  #include <rte_ether.h>
+#include <rte_gtp.h>
  #include <rte_icmp.h>
  #include <rte_ip.h>
  #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
   *
   * Matches a GTP PDU extension header with type 0x85.
   */
+RTE_STD_C11
  struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1439

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +9 -0
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
  };

This will change the struct size even with union, since old version is missing all fields from protocol header. Struct size will go from 2 --> 5 bytes [1].

Since this is public struct, we can't change its size.

@Ori, Andrew,

Do you have a suggestion for next step?

- We can still add the "struct rte_gtp_psc_hdr", and add a deprecation notice for "struct rte_flow_item_gtp_psc" to say it will use "struct rte_gtp_psc_hdr" on 21.11.

- And for this release use the Raslan's first version:
- uint8_t qfi; /**< QoS flow identifier. */
+ uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */

Does it make sense? Any comments?

[1]

struct rte_gtp_psc_hdr {
        uint8_t                    ext_hdr_len;          /*     0     1 */
        uint8_t                    type:4;               /*     1: 0  1 */
        uint8_t                    qmp:1;                /*     1: 4  1 */

        /* XXX 3 bits hole, try to pack */

        union {
                struct {
                        uint8_t    snp:1;                /*     2: 0  1 */
                        uint8_t    spare_dl1:2;          /*     2: 1  1 */
                };                                       /*     2     1 */
                struct {
                        uint8_t    dl_delay_ind:1;       /*     2: 0  1 */
                        uint8_t    ul_delay_ind:1;       /*     2: 1  1 */
                        uint8_t    snp_ul1:1;            /*     2: 2  1 */
                };                                       /*     2     1 */
        };                                               /*     2     1 */
        union {
                struct {
                        uint8_t    ppp:1;                /*     3: 0  1 */
                        uint8_t    rqi:1;                /*     3: 1  1 */
                };                                       /*     3     1 */
                struct {
                        uint8_t    n_delay_ind:1;        /*     3: 0  1 */
                        uint8_t    spare_ul2:1;          /*     3: 1  1 */
                };                                       /*     3     1 */
        };                                               /*     3     1 */
        uint8_t                    qfi:6;                /*     4: 0  1 */

        /* XXX 2 bits hole, try to pack */

        uint8_t                    data[];               /*     5     0 */

        /* size: 5, cachelines: 1, members: 7 */
        /* sum members: 3 */
        /* sum bitfield members: 11 bits, bit holes: 2, sum bit holes: 5 bits */
        /* last cacheline: 5 bytes */
};
Quoted 13 lines
lib/librte_ethdev/rte_flow.h +3 -1
  /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
  #ifndef __cplusplus
  static const struct rte_flow_item_gtp_psc
  rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
  };
  #endif
 
 
#25 lore raw

Hi,

Quoted 99 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Tuesday, April 6, 2021 7:10 PM
To: Raslan Darawsheh <[email protected]>; [email protected]; Ori Kam <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; David Marchand <[email protected]>; NBU-Contact-Thomas Monjalon <[email protected]>
Subject: Re: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 15 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
  lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
  2 files changed, 19 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 22 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +1 -0
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
  #include <rte_arp.h>
  #include <rte_common.h>
  #include <rte_ether.h>
+#include <rte_gtp.h>
  #include <rte_icmp.h>
  #include <rte_ip.h>
  #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +4 -2
   *
   * Matches a GTP PDU extension header with type 0x85.
   */
+RTE_STD_C11
  struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1439

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +9 -0
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
  };

This will change the struct size even with union, since old version is missing all fields from protocol header. Struct size will go from 2 --> 5 bytes [1].

Since this is public struct, we can't change its size.

@Ori, Andrew,

Do you have a suggestion for next step?

- We can still add the "struct rte_gtp_psc_hdr", and add a deprecation notice for "struct rte_flow_item_gtp_psc" to say it will use "struct rte_gtp_psc_hdr"
on 21.11.

Quoted 7 lines

- And for this release use the Raslan's first version:
- uint8_t qfi; /**< QoS flow identifier. */
+ uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */

Does it make sense? Any comments?

@Ori Kam, @[email protected],
This is a kind remainder of this patch,

Quoted 56 lines

[1]

struct rte_gtp_psc_hdr {
         uint8_t                    ext_hdr_len;          /*     0     1 */
         uint8_t                    type:4;               /*     1: 0  1 */
         uint8_t                    qmp:1;                /*     1: 4  1 */

         /* XXX 3 bits hole, try to pack */

         union {
                 struct {
                         uint8_t    snp:1;                /*     2: 0  1 */
                         uint8_t    spare_dl1:2;          /*     2: 1  1 */
                 };                                       /*     2     1 */
                 struct {
                         uint8_t    dl_delay_ind:1;       /*     2: 0  1 */
                         uint8_t    ul_delay_ind:1;       /*     2: 1  1 */
                         uint8_t    snp_ul1:1;            /*     2: 2  1 */
                 };                                       /*     2     1 */
         };                                               /*     2     1 */
         union {
                 struct {
                         uint8_t    ppp:1;                /*     3: 0  1 */
                         uint8_t    rqi:1;                /*     3: 1  1 */
                 };                                       /*     3     1 */
                 struct {
                         uint8_t    n_delay_ind:1;        /*     3: 0  1 */
                         uint8_t    spare_ul2:1;          /*     3: 1  1 */
                 };                                       /*     3     1 */
         };                                               /*     3     1 */
         uint8_t                    qfi:6;                /*     4: 0  1 */

         /* XXX 2 bits hole, try to pack */

         uint8_t                    data[];               /*     5     0 */

         /* size: 5, cachelines: 1, members: 7 */
         /* sum members: 3 */
         /* sum bitfield members: 11 bits, bit holes: 2, sum bit holes: 5 bits */
         /* last cacheline: 5 bytes */
};
Quoted 13 lines
lib/librte_ethdev/rte_flow.h +3 -1
  /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
  #ifndef __cplusplus
  static const struct rte_flow_item_gtp_psc
  rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
  };
  #endif
 
 
#26 lore raw

Hi Raslan,

Quoted 4 lines

-----Original Message-----
From: Raslan Darawsheh <[email protected]>
Subject: RE: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 96 lines

Hi,

Quoted 94 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Tuesday, April 6, 2021 7:10 PM
To: Raslan Darawsheh <[email protected]>; [email protected]; Ori Kam <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; David Marchand <[email protected]>; NBU-Contact-Thomas Monjalon <[email protected]>
Subject: Re: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 15 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
  lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
  2 files changed, 19 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 22 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
  - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
    - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
  - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +1 -0
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
  #include <rte_arp.h>
  #include <rte_common.h>
  #include <rte_ether.h>
+#include <rte_gtp.h>
  #include <rte_icmp.h>
  #include <rte_ip.h>
  #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +4 -2
   *
   * Matches a GTP PDU extension header with type 0x85.
   */
+RTE_STD_C11
  struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1439

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +9 -0
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
  };

This will change the struct size even with union, since old version is missing all fields from protocol header. Struct size will go from 2 --> 5 bytes [1].

Since this is public struct, we can't change its size.

@Ori, Andrew,

Do you have a suggestion for next step?

I think Ferruh is right, and I think that we should at this point just update the documentation. Sorry for the detour
Just small comment about the original patch.
I don’t think you should change the default mask since it may break existing apps.

Quoted 71 lines
Quoted 5 lines

- We can still add the "struct rte_gtp_psc_hdr", and add a deprecation notice for "struct rte_flow_item_gtp_psc" to say it will use "struct rte_gtp_psc_hdr"
on 21.11.

Quoted 7 lines

- And for this release use the Raslan's first version:
- uint8_t qfi; /**< QoS flow identifier. */
+ uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */

Does it make sense? Any comments?

@Ori Kam, @[email protected],
This is a kind remainder of this patch,

Quoted 56 lines

[1]

struct rte_gtp_psc_hdr {
         uint8_t                    ext_hdr_len;          /*     0     1 */
         uint8_t                    type:4;               /*     1: 0  1 */
         uint8_t                    qmp:1;                /*     1: 4  1 */

         /* XXX 3 bits hole, try to pack */

         union {
                 struct {
                         uint8_t    snp:1;                /*     2: 0  1 */
                         uint8_t    spare_dl1:2;          /*     2: 1  1 */
                 };                                       /*     2     1 */
                 struct {
                         uint8_t    dl_delay_ind:1;       /*     2: 0  1 */
                         uint8_t    ul_delay_ind:1;       /*     2: 1  1 */
                         uint8_t    snp_ul1:1;            /*     2: 2  1 */
                 };                                       /*     2     1 */
         };                                               /*     2     1 */
         union {
                 struct {
                         uint8_t    ppp:1;                /*     3: 0  1 */
                         uint8_t    rqi:1;                /*     3: 1  1 */
                 };                                       /*     3     1 */
                 struct {
                         uint8_t    n_delay_ind:1;        /*     3: 0  1 */
                         uint8_t    spare_ul2:1;          /*     3: 1  1 */
                 };                                       /*     3     1 */
         };                                               /*     3     1 */
         uint8_t                    qfi:6;                /*     4: 0  1 */

         /* XXX 2 bits hole, try to pack */

         uint8_t                    data[];               /*     5     0 */

         /* size: 5, cachelines: 1, members: 7 */
         /* sum members: 3 */
         /* sum bitfield members: 11 bits, bit holes: 2, sum bit holes: 5 bits */
         /* last cacheline: 5 bytes */
};
Quoted 13 lines
lib/librte_ethdev/rte_flow.h +3 -1
  /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
  #ifndef __cplusplus
  static const struct rte_flow_item_gtp_psc
  rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
  };
  #endif
 
 
#27 lore raw
Quoted 100 lines · Ori Kam

Hi Raslan,

Quoted 98 lines

-----Original Message-----
From: Raslan Darawsheh <[email protected]>
Subject: RE: [PATCH v4 2/2] ethdev: update qfi definition

Hi,

Quoted 92 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Tuesday, April 6, 2021 7:10 PM
To: Raslan Darawsheh <[email protected]>; [email protected]; Ori Kam <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; David Marchand <[email protected]>; NBU-Contact-Thomas Monjalon <[email protected]>
Subject: Re: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 15 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
   doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
   lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
   2 files changed, 19 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 22 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
   - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
     - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
   - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +1 -0
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
   #include <rte_arp.h>
   #include <rte_common.h>
   #include <rte_ether.h>
+#include <rte_gtp.h>
   #include <rte_icmp.h>
   #include <rte_ip.h>
   #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +4 -2
    *
    * Matches a GTP PDU extension header with type 0x85.
    */
+RTE_STD_C11
   struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1439

Quoted 10 lines
lib/librte_ethdev/rte_flow.h +9 -0
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
   };

This will change the struct size even with union, since old version is missing all fields from protocol header. Struct size will go from 2 --> 5 bytes [1].

Since this is public struct, we can't change its size.

@Ori, Andrew,

Quoted 8 lines
Quoted 2 lines
Quoted 2 lines

Do you have a suggestion for next step?

I think Ferruh is right, and I think that we should at this point just update the documentation. Sorry for the detour
Just small comment about the original patch.
I don’t think you should change the default mask since it may break existing apps.

I will continue with first patch [1], and will update the protocol reference in the commit log.

Meanwhile adding 'struct rte_gtp_psc_hdr' (patch v4 1/2) work can continue separately.
And when it is added we can send a deprecation notice to update 'struct rte_flow_item_gtp_psc' and finally update it on v21.11 to have 'struct rte_gtp_psc_hdr' in it. Makes sense?

[1]
https://patches.dpdk.org/project/dpdk/patch/[email protected]/

Quoted 72 lines
Quoted 71 lines
Quoted 5 lines

- We can still add the "struct rte_gtp_psc_hdr", and add a deprecation notice for "struct rte_flow_item_gtp_psc" to say it will use "struct rte_gtp_psc_hdr"
on 21.11.

Quoted 7 lines

- And for this release use the Raslan's first version:

    -	uint8_t qfi; /**< QoS flow identifier. */
    +	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */

Does it make sense? Any comments?

@Ori Kam, @[email protected],
This is a kind remainder of this patch,

Quoted 56 lines

[1]

struct rte_gtp_psc_hdr {
          uint8_t                    ext_hdr_len;          /*     0     1 */
          uint8_t                    type:4;               /*     1: 0  1 */
          uint8_t                    qmp:1;                /*     1: 4  1 */

          /* XXX 3 bits hole, try to pack */

          union {
                  struct {
                          uint8_t    snp:1;                /*     2: 0  1 */
                          uint8_t    spare_dl1:2;          /*     2: 1  1 */
                  };                                       /*     2     1 */
                  struct {
                          uint8_t    dl_delay_ind:1;       /*     2: 0  1 */
                          uint8_t    ul_delay_ind:1;       /*     2: 1  1 */
                          uint8_t    snp_ul1:1;            /*     2: 2  1 */
                  };                                       /*     2     1 */
          };                                               /*     2     1 */
          union {
                  struct {
                          uint8_t    ppp:1;                /*     3: 0  1 */
                          uint8_t    rqi:1;                /*     3: 1  1 */
                  };                                       /*     3     1 */
                  struct {
                          uint8_t    n_delay_ind:1;        /*     3: 0  1 */
                          uint8_t    spare_ul2:1;          /*     3: 1  1 */
                  };                                       /*     3     1 */
          };                                               /*     3     1 */
          uint8_t                    qfi:6;                /*     4: 0  1 */

          /* XXX 2 bits hole, try to pack */

          uint8_t                    data[];               /*     5     0 */

          /* size: 5, cachelines: 1, members: 7 */
          /* sum members: 3 */
          /* sum bitfield members: 11 bits, bit holes: 2, sum bit holes: 5 bits */
          /* last cacheline: 5 bytes */
};
Quoted 13 lines
lib/librte_ethdev/rte_flow.h +3 -1
   /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
   #ifndef __cplusplus
   static const struct rte_flow_item_gtp_psc
   rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
   };
   #endif
 
 
#28 lore raw

Kindest regards,
Raslan Darawsheh

Quoted 126 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, April 14, 2021 3:17 PM
To: Ori Kam <[email protected]>; Raslan Darawsheh <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>; [email protected]; David Marchand <[email protected]>; NBU-Contact-Thomas Monjalon <[email protected]>
Subject: Re: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 15 lines · Ori Kam

Hi Raslan,

Quoted 13 lines

-----Original Message-----
From: Raslan Darawsheh <[email protected]>
Subject: RE: [PATCH v4 2/2] ethdev: update qfi definition

Hi,

Quoted 7 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Tuesday, April 6, 2021 7:10 PM
To: Raslan Darawsheh <[email protected]>; [email protected]; Ori Kam <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>;

Shiri

Quoted 71 lines
Quoted 71 lines
Quoted 71 lines

Kuzin <[email protected]>; [email protected]; David Marchand <[email protected]>; NBU-Contact-Thomas Monjalon <[email protected]>
Subject: Re: [PATCH v4 2/2] ethdev: update qfi definition

Quoted 15 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Review

Quoted 7 lines
---
   doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
   lib/librte_ethdev/rte_flow.h                | 20 +++++++++++++++++---
   2 files changed, 19 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/doc/guides/testpmd_app_ug/testpmd_funcs.rst

doc/guides/testpmd_app_ug/testpmd_funcs.rst This section lists supported pattern items and

Quoted 4 lines

index f59eb8a27d..dd39c4c3c2 100644

doc/guides/testpmd_app_ug/testpmd_funcs.rst +0 -0
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and

their attributes, if any.

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta

Quoted 22 lines
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
   - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
     - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
   - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +1 -0
index 6cc57136ac..20b4389429 100644
@@ -20,6 +20,7 @@
   #include <rte_arp.h>
   #include <rte_common.h>
   #include <rte_ether.h>
+#include <rte_gtp.h>
   #include <rte_icmp.h>
   #include <rte_ip.h>
   #include <rte_sctp.h>
@@ -1421,16 +1422,29 @@ static const struct rte_flow_item_meta

rte_flow_item_meta_mask = {

Quoted 14 lines
lib/librte_ethdev/rte_flow.h +8 -2
    *
    * Matches a GTP PDU extension header with type 0x85.
    */
+RTE_STD_C11
   struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

lib/librte_ethdev/rte_flow.h static const struct rte_flow_item_meta rte_flow_item_meta_mask = { · line 1439

Quoted 6 lines
lib/librte_ethdev/rte_flow.h +5 -0
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc_hdr hdr;
+	};
   };

This will change the struct size even with union, since old version is

missing

Quoted 10 lines
Quoted 8 lines
Quoted 8 lines

all fields from protocol header. Struct size will go from 2 --> 5 bytes [1].

Since this is public struct, we can't change its size.

@Ori, Andrew,

Do you have a suggestion for next step?

I think Ferruh is right, and I think that we should at this point just update

the documentation.

Quoted 3 lines

Sorry for the detour
Just small comment about the original patch.
I don't think you should change the default mask since it may break existing

apps.

Quoted 1 line
Quoted 9 lines

I will continue with first patch [1], and will update the protocol reference in the commit log.

Meanwhile adding 'struct rte_gtp_psc_hdr' (patch v4 1/2) work can continue separately.
And when it is added we can send a deprecation notice to update 'struct rte_flow_item_gtp_psc' and finally update it on v21.11 to have 'struct rte_gtp_psc_hdr' in it. Makes sense?

Yes, makes sense to me,
I'll work in the meanwhile on providing the new hdrs definition .

Quoted 78 lines

[1]
https://patches.dpdk.org/project/dpdk/patch/20210323121134.19113-1-
[email protected]/

Quoted 1 line
Quoted 1 line
Quoted 1 line

- We can still add the "struct rte_gtp_psc_hdr", and add a deprecation

notice

Quoted 47 lines
Quoted 47 lines
Quoted 11 lines

for "struct rte_flow_item_gtp_psc" to say it will use "struct rte_gtp_psc_hdr"
on 21.11.

- And for this release use the Raslan's first version:

    -	uint8_t qfi; /**< QoS flow identifier. */
    +	uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */

Does it make sense? Any comments?

@Ori Kam, @[email protected],
This is a kind remainder of this patch,

Quoted 33 lines

[1]

struct rte_gtp_psc_hdr {
          uint8_t                    ext_hdr_len;          /*     0     1 */
          uint8_t                    type:4;               /*     1: 0  1 */
          uint8_t                    qmp:1;                /*     1: 4  1 */

          /* XXX 3 bits hole, try to pack */

          union {
                  struct {
                          uint8_t    snp:1;                /*     2: 0  1 */
                          uint8_t    spare_dl1:2;          /*     2: 1  1 */
                  };                                       /*     2     1 */
                  struct {
                          uint8_t    dl_delay_ind:1;       /*     2: 0  1 */
                          uint8_t    ul_delay_ind:1;       /*     2: 1  1 */
                          uint8_t    snp_ul1:1;            /*     2: 2  1 */
                  };                                       /*     2     1 */
          };                                               /*     2     1 */
          union {
                  struct {
                          uint8_t    ppp:1;                /*     3: 0  1 */
                          uint8_t    rqi:1;                /*     3: 1  1 */
                  };                                       /*     3     1 */
                  struct {
                          uint8_t    n_delay_ind:1;        /*     3: 0  1 */
                          uint8_t    spare_ul2:1;          /*     3: 1  1 */
                  };                                       /*     3     1 */
          };                                               /*     3     1 */
          uint8_t                    qfi:6;                /*     4: 0  1 */
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
          /* XXX 2 bits hole, try to pack */

          uint8_t                    data[];               /*     5     0 */

          /* size: 5, cachelines: 1, members: 7 */
          /* sum members: 3 */
          /* sum bitfield members: 11 bits, bit holes: 2, sum bit holes: 5 bits

*/

Quoted 17 lines
Quoted 16 lines
Quoted 16 lines
          /* last cacheline: 5 bytes */
};
Quoted 13 lines
lib/librte_ethdev/rte_flow.h +3 -1
   /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
   #ifndef __cplusplus
   static const struct rte_flow_item_gtp_psc
   rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.hdr.ppp = 1,
+	.hdr.rqi = 1,
+	.hdr.qfi = 0x3f,
   };
   #endif
 
 
#29 lore raw

v5 cover add new hdr for gtp qfi

In reply to Raslan Darawsheh

This is introducin a new hdr definition of gtp psc support to match the RFC 38415-g30

v2: introduce new header definition for gtp psc
update commit msg for rte flow item change.

v3: fixed typo in comment
Cc relevant people.

v4: update hdr definition to have hdr suffix.
update variable name to be hdr in the gtp_psc item. update default max to use the new added hdr.

v5: updated the hdr definition after code review.
dropped the change for rte_flow item psc to avoid ABI breakage added hdr definition for type0 and type1 psc's

Raslan Darawsheh (1):
ethdev: add new ext hdr for gtp psc

 lib/net/rte_gtp.h | 78 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
Signature · 1 line
2.25.1
#30 lore raw

PATCH v5 1/1 ethdev: add new ext hdr for gtp psc

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/net/rte_gtp.h | 78 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
lib/net/rte_gtp.h +78 -0
index 6a6f9b238d..5a850a26e4 100644
@@ -61,6 +61,84 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_generic_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type0 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type0_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type1 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type1_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
Signature · 1 line
2.25.1
#31 lore raw

Re: [dpdk-dev] [PATCH v5 1/1] ethdev: add new ext hdr for gtp psc

Quoted 4 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

This patch has not urgent for this release, postponing it to next release.

#32 lore raw

Re: [dpdk-dev] [PATCH v5 1/1] ethdev: add new ext hdr for gtp psc

In reply to Raslan Darawsheh
Quoted 88 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/net/rte_gtp.h | 78 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
lib/net/rte_gtp.h +71 -0
index 6a6f9b238d..5a850a26e4 100644
@@ -61,6 +61,84 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_generic_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type0 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type0_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type1 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type1_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */

lib/net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 144

Quoted 10 lines
lib/net/rte_gtp.h +7 -0
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))

This way the structure is very hard to read.
May I ask to indent field comments as it is done in other rte_net structures (e.g. rte_ipv4_hdr, rte_udp_hdr, rte_geneve_hdr). Personally I don't care if it will be indent by TABs or spaces, however, it looks like TABs are used in more places.

#33 lore raw

Re: [dpdk-dev] [PATCH v5 1/1] ethdev: add new ext hdr for gtp psc

Quoted 98 lines · Andrew Rybchenko
Quoted 97 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/net/rte_gtp.h | 78 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
lib/net/rte_gtp.h +78 -0
index 6a6f9b238d..5a850a26e4 100644
@@ -61,6 +61,84 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_generic_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type0 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type0_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type1 defined based on RFC 38415-g30
+ */
+__extension__
+struct rte_gtp_psc_type1_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes */
+#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+#else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */
+} __rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))
Quoted 9 lines
Quoted 1 line

This way the structure is very hard to read.
May I ask to indent field comments as it is done in other rte_net structures (e.g. rte_ipv4_hdr, rte_udp_hdr, rte_geneve_hdr). Personally I don't care if it will be indent by TABs or spaces, however, it looks like TABs are used in more places.

Also there are spelling bugs [1].

[1] http://mails.dpdk.org/archives/test-report/2021-April/191649.html

#34 lore raw

Re: [dpdk-dev] [PATCH v5 1/1] ethdev: add new ext hdr for gtp psc

Hi Andrew,

Thanks for your comments and review, I'll attempt to handle and fix in the next version.

Kindest regards,
Raslan Darawsheh

Quoted 114 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Tuesday, June 8, 2021 3:19 PM
To: Raslan Darawsheh <[email protected]>; [email protected]
Cc: [email protected]; Ori Kam <[email protected]>; [email protected]; [email protected]; [email protected]; Slava Ovsiienko <[email protected]>; Shiri Kuzin <[email protected]>
Subject: Re: [dpdk-dev] [PATCH v5 1/1] ethdev: add new ext hdr for gtp psc

Quoted 91 lines · Andrew Rybchenko
Quoted 90 lines · Raslan Darawsheh

Define new rte header for gtp PDU session container based on RFC 38415-g30

---
 lib/net/rte_gtp.h | 78
+++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
lib/net/rte_gtp.h index renamed from lib/net/rte_gtp.h +0 -0

6a6f9b238d..5a850a26e4 100644

lib/net/rte_gtp.h +71 -0
@@ -61,6 +61,84 @@ struct rte_gtp_hdr_ext_word {
 	uint8_t next_ext;     /**< Next Extension Header Type. */
 }  __rte_packed;
 
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * defined based on RFC 38415-g30.
+ */
+__extension__
+struct rte_gtp_psc_generic_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes
+*/ #if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */ #else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare:2; /**< type specific spare bits */
+	uint8_t pad:3; /**< type specfic pad bits */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */ }
+__rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type0 defined based on RFC 38415-g30  */ __extension__ struct
+rte_gtp_psc_type0_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes
+*/ #if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t qfi:6; /**< Qos Flow Identifier */ #else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t rqi:1; /**< Reflective Qos Indicator */
+	uint8_t ppp:1; /**< Paging policy presence */
+	uint8_t spare_dl1:2; /**< spare down link bits */
+	uint8_t snp:1; /**< Sequence number presence */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */ }
+__rte_packed;
+
+/**
+ * Optional extension for GTP with next_ext set to 0x85
+ * type1 defined based on RFC 38415-g30  */ __extension__ struct
+rte_gtp_psc_type1_hdr {
+	uint8_t ext_hdr_len; /**< PDU ext hdr len in multiples of 4 bytes
+*/ #if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
+	uint8_t type:4; /**< PDU type */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t qfi:6; /**< Qos Flow Identifier */ #else
+	uint8_t qfi:6; /**< Qos Flow Identifier */
+	uint8_t spare_ul2:1; /**< spare up link bits */
+	uint8_t n_delay_ind:1; /**< N3/N9 delay result presence */
+	uint8_t snp:1; /**< Sequence number presence ul */
+	uint8_t ul_delay_ind:1; /**< ul delay result presence */
+	uint8_t dl_delay_ind:1; /**< dl delay result presence */
+	uint8_t qmp:1; /**< Qos Monitoring Packet */
+	uint8_t type:4; /**< PDU type */
+#endif
+	uint8_t data[0]; /**< variable length data feilds */ }
+__rte_packed;
+
 /** GTP header length */
 #define RTE_ETHER_GTP_HLEN \
Quoted 9 lines

lib/net/rte_gtp.h struct rte_gtp_hdr_ext_word { · line 144

Quoted 2 lines
lib/net/rte_gtp.h +0 -0
 	(sizeof(struct rte_udp_hdr) + sizeof(struct rte_gtp_hdr))

This way the structure is very hard to read.
May I ask to indent field comments as it is done in other rte_net structures (e.g. rte_ipv4_hdr, rte_udp_hdr, rte_geneve_hdr). Personally I don't care if it will be indent by TABs or spaces, however, it looks like TABs are used in more places.

Also there are spelling bugs [1].

[1] http://mails.dpdk.org/archives/test-report/2021-April/191649.html

#35 lore raw

PATCH v2 2/2 ethdev: update qfi definition

In reply to Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier) based on RFC 38415-g30

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")
---
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  3 ++-
 lib/librte_ethdev/rte_flow.h                | 18 +++++++++++++++---
 2 files changed, 17 insertions(+), 4 deletions(-)
doc/guides/testpmd_app_ug/testpmd_funcs.rst +2 -1
index f59eb8a27d..dd39c4c3c2 100644
@@ -3742,7 +3742,8 @@ This section lists supported pattern items and their attributes, if any.
 - ``gtp_psc``: match GTP PDU extension header with type 0x85.
 
   - ``pdu_type {unsigned}``: PDU type.
-  - ``qfi {unsigned}``: QoS flow identifier.
+
+  - ``qfi {unsigned}``: PPP, RQI and QoS flow identifier.
 
 - ``pppoes``, ``pppoed``: match PPPoE header.
 
lib/librte_ethdev/rte_flow.h +15 -3
index 6cc57136ac..1eb9711707 100644
@@ -20,6 +20,7 @@
 #include <rte_arp.h>
 #include <rte_common.h>
 #include <rte_ether.h>
+#include <rte_gtp.h>
 #include <rte_icmp.h>
 #include <rte_ip.h>
 #include <rte_sctp.h>
@@ -1421,16 +1422,27 @@ static const struct rte_flow_item_meta rte_flow_item_meta_mask = {
  *
  * Matches a GTP PDU extension header with type 0x85.
  */
+RTE_STD_C11
 struct rte_flow_item_gtp_psc {
-	uint8_t pdu_type; /**< PDU type. */
-	uint8_t qfi; /**< QoS flow identifier. */
+	union {
+		struct {
+			/*
+			 * These fields are retained for compatibility.
+			 * Please switch to the new header field below.
+			 */
+			uint8_t pdu_type; /**< PDU type. */
+			uint8_t qfi; /**< PPP, RQI, QoS flow identifier. */
+
+		};
+		struct rte_gtp_psc gtp_psc;
+	};
 };
 
 /** Default mask for RTE_FLOW_ITEM_TYPE_GTP_PSC. */
 #ifndef __cplusplus
 static const struct rte_flow_item_gtp_psc
 rte_flow_item_gtp_psc_mask = {
-	.qfi = 0x3f,
+	.qfi = 0xff,
 };
 #endif
 
Signature · 1 line
2.29.0
Reviewed-by on v1 1/1 Applied
#36 lore raw
In reply to Raslan Darawsheh
Quoted 8 lines · Raslan Darawsheh

qfi field is 8 bits which represent single bit for PPP (paging Policy Presence) single bit for RQI (Reflective QoS Indicator) and 6 bits for qfi (QoS Flow Identifier)

This update the doxygen format and the mask for qfi to properly identify the full 8 bits of the field.

Quoted 8 lines

note: changing the default mask would cause different patterns generated by testpmd.

Fixes 346553db5bd1 ("ethdev: add GTP extension header to flow API")

Applied to dpdk-next-net/main, thanks.

Sorry for going through multiple versions, hopefully proper solution can be merged after deprecation notice on v21.11.