Skip to contentSkip to conversation

doc: announce header split deprecation

PATCH Superseded [email protected] 6 participants 20 messages May 23, 2022 – Jul 16, 2022

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 doc: announce header split deprecation

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +4 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the ``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure ``rte_eth_rxmode`` to enable header
+split offload are not supported in any PMDs. They will be removed in DPDK 22.11.
Signature · 1 line
2.17.1
Acked-by on v1 1/1
#1 lore raw
Quoted 12 lines · [email protected] writes:

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

Quoted 4 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

#2 lore raw
Quoted 15 lines · Ray Kinsella
Quoted 12 lines · [email protected] writes:

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

Quoted 7 lines
Quoted 4 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

#3 lore raw
In reply to [email protected]

23/05/2022 16:20, [email protected]:

Quoted 25 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 131

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +4 -0
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the ``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure ``rte_eth_rxmode`` to enable header
+split offload are not supported in any PMDs. They will be removed in DPDK 22.11.

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

author
#4 lore raw

Hi,

Quoted 47 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: 2022年7月13日 17:50
To: Ding, Xuan <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time

ago

Quoted 8 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 15 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +5 -0
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol based buffer split? About the deprecation of header split, I haven't realized its connection to rte_eth_rxseg_split.

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

Regards,
Xuan

Quoted 1 line
#5 lore raw

14/07/2022 07:50, Ding, Xuan:

Quoted 40 lines

From: Thomas Monjalon <[email protected]>

Quoted 39 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time

ago

Quoted 8 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 15 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +5 -0
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol based buffer split? About the deprecation of header split, I haven't realized its connection to rte_eth_rxseg_split.

What???
In old versions of your patch "ethdev: introduce protocol type based header split" you wrote:
"
A new proto field is introduced in the
rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

author
#6 lore raw

Hi,

Quoted 65 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 4:09 PM
To: Ding, Xuan <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 24 lines
Quoted 24 lines
Quoted 13 lines

will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 9 lines
doc/guides/rel_notes/deprecation.rst +6 -0
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 6 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 1 line

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 10 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the
rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using "split_hdr_size". But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT for this purpose. So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in rte_eth_rxmode to configure split location.

Quoted 6 lines

"

Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But we can still clean the code. Hope it make things clearer.

Thanks,
Xuan

Quoted 1 line
#7 lore raw

14/07/2022 14:54, Ding, Xuan:

Quoted 76 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 55 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 24 lines
Quoted 24 lines
Quoted 13 lines

will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 9 lines
doc/guides/rel_notes/deprecation.rst +6 -0
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 6 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 1 line

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the
rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using "split_hdr_size". But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT for this purpose. So we finally decide to deprecate this flag.

Quoted 4 lines

http://patchwork.dpdk.org/project/dpdk/patch/[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But we can still clean the code. Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue,
while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

author
#8 lore raw

Hi,

Quoted 112 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 22 lines
Quoted 22 lines

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
 doc/guides/rel_notes/deprecation.rst | 4 ++++
 1 file changed, 4 insertions(+)
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 125

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library

instead,

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 126

Quoted 1 line
doc/guides/rel_notes/deprecation.rst +0 -0
   with the underlying HW-functionality being provided by the ``ioat``

or

Quoted 33 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +6 -0
   ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type.
With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 3 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line

But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-w

Quoted 4 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 8 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Quoted 3 lines

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

Agree. It is better to point out the remaining per queue rx offload RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT for splitting packets.

Please see v2 after I add more header split background.

Regards,
Xuan

Quoted 1 line
#9 lore raw
Quoted 123 lines · Xuan

Hi,

Quoted 112 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 22 lines
Quoted 22 lines

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++
  1 file changed, 4 insertions(+)
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 125

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
    applications should be updated to use the ``dmadev`` library

instead,

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 126

Quoted 1 line
doc/guides/rel_notes/deprecation.rst +0 -0
    with the underlying HW-functionality being provided by the ``ioat``

or

Quoted 33 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +6 -0
    ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type.
With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 3 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line

But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-w

Quoted 4 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 8 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Quoted 4 lines

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

As far as I understand 'BUFFER_SPLIT' has a target to split package based on size, to various size mempools, and it can split the payload as well without caring the protocol headers. Most probably practical usage is to split inner protocol, but protocol not needed to be known by NIC, it can still split based on size.

Also as far as I can see mlx5 implemented the feature [1], contradicting the claim in the commit log, again Slava can comment better. (Although I don't see any PMD claim "Buffer Split on Rx" feature supported.)

[1]
Commit 9f209b59c8b0 ("net/mlx5: support Rx buffer split description")

Quoted 9 lines
Quoted 3 lines

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

Agree. It is better to point out the remaining per queue rx offload RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT for splitting packets.

Please see v2 after I add more header split background.

I can't see a v2 in patchwork, is a v2 sent? If so, may be something went wrong.

Acked-by
#10 lore raw
Quoted 129 lines · Ferruh Yigit
Quoted 123 lines · Xuan

Hi,

Quoted 112 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 22 lines
Quoted 22 lines

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++

  1 file changed, 4 insertions(+)
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices

    applications should be updated to use the ``dmadev`` library

instead,

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

    with the underlying HW-functionality being provided by the ``ioat``

or

Quoted 33 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

    ``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type.
With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 3 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line

But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-w

Quoted 4 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 8 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Quoted 4 lines

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Quoted 6 lines

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

As far as I understand 'BUFFER_SPLIT' has a target to split package based on size, to various size mempools, and it can split the payload as well without caring the protocol headers. Most probably practical usage is to split inner protocol, but protocol not needed to be known by NIC, it can still split based on size.

Ahh, deprecation notice is to deprecate 'HEADER_SPLIT', not other way around as I got it wrongly.

I think both may have use case, specially for NICs that parse the protocols, header split can be easier to use. But since there is no user for a long time, perhaps that is not a real life need, and we can always add it back when needed, hence:

Quoted 19 lines

Also as far as I can see mlx5 implemented the feature [1], contradicting the claim in the commit log, again Slava can comment better. (Although I don't see any PMD claim "Buffer Split on Rx" feature supported.)

[1]
Commit 9f209b59c8b0 ("net/mlx5: support Rx buffer split description")

Quoted 9 lines
Quoted 3 lines

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

Agree. It is better to point out the remaining per queue rx offload RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT for splitting packets.

Please see v2 after I add more header split background.

I can't see a v2 in patchwork, is a v2 sent? If so, may be something went wrong.

author
#11 lore raw

Hi Ferruh,

Quoted 158 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, July 15, 2022 12:03 AM
To: Ding, Xuan <[email protected]>; Thomas Monjalon <[email protected]>; [email protected]; Viacheslav Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 92 lines · Ferruh Yigit
Quoted 91 lines · Xuan

Hi,

Quoted 89 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 22 lines
Quoted 22 lines

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++

  1 file changed, 4 insertions(+)
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices

    applications should be updated to use the ``dmadev`` library

instead,

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

    with the underlying HW-functionality being provided by the ``ioat``

or

Quoted 33 lines
Quoted 32 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

    ``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line

But no single PMD actually supports

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

Quoted 40 lines
Quoted 31 lines
Quoted 22 lines

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-w

Quoted 5 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 8 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Quoted 4 lines

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

Quoted 6 lines

As far as I understand 'BUFFER_SPLIT' has a target to split package based on size, to various size mempools, and it can split the payload as well without caring the protocol headers.
Most probably practical usage is to split inner protocol, but protocol not needed to be known by NIC, it can still split based on size.

Quoted 5 lines

Ahh, deprecation notice is to deprecate 'HEADER_SPLIT', not other way around as I got it wrongly.

I think both may have use case, specially for NICs that parse the protocols, header split can be easier to use.

Thanks for your comments.

Does it supports to split packets into different mempools?
Actually in DPDK22.07, we send another series to extend current buffer split to support split by protocol header.

http://patchwork.dpdk.org/project/dpdk/cover/[email protected]/

Since no PMD supports 'HEADER_SPLIT', I cannot find a reference.
IMO the 'HEADER SPLIT' is not used for splitting by protocol headers, it still requires 'split_hdr_size'.

Quoted 11 lines

But since there is no user for a long time, perhaps that is not a real life need, and we can always add it back when needed, hence:

Quoted 5 lines

Also as far as I can see mlx5 implemented the feature [1], contradicting the claim in the commit log, again Slava can comment better. (Although I don't see any PMD claim "Buffer Split on Rx" feature supported.)

Yes, you are right.

Quoted 9 lines
Quoted 9 lines

[1]
Commit 9f209b59c8b0 ("net/mlx5: support Rx buffer split description")

Quoted 4 lines
Quoted 3 lines

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Agree. It is better to point out the remaining per queue rx offload RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT for splitting packets.

Please see v2 after I add more header split background.

I can't see a v2 in patchwork, is a v2 sent? If so, may be something went wrong.

Sorry, the v2 has not been sent yet since there are doubts about the relationship header split and buffer split.

Please see last email. Hope to get your insights.

Regards,
Xuan

Quoted 1 line
Quoted 1 line
#12 lore raw
In reply to Ferruh Yigit
Quoted 138 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, July 14, 2022 18:58
To: Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected]; Slava Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 126 lines · Xuan

Hi,

Quoted 123 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 22 lines
Quoted 22 lines

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++
  1 file changed, 4 insertions(+)
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 125

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
    applications should be updated to use the ``dmadev`` library

instead,

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
doc/guides/rel_notes/deprecation.rst +0 -0
    with the underlying HW-functionality being provided by the

``ioat``

or

Quoted 33 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 3 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line

But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-w

Quoted 5 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Quoted 8 lines
Quoted 4 lines

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

Hi, thank you for Cc'ing

Yes, you are right, we have two splitting offloads, and these ones have the different intentions. As there are no PMDs actually handling RTE_ETH_RX_OFFLOAD_HEADER_SPLIT, there should be no objections for this deprecation.

author
#13 lore raw

Hi,

Quoted 168 lines

-----Original Message-----
From: Slava Ovsiienko <[email protected]>
Sent: Friday, July 15, 2022 12:56 AM
To: Ferruh Yigit <[email protected]>; Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: RE: [PATCH] doc: announce header split deprecation

Quoted 5 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, July 14, 2022 18:58
To: Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected];

Slava

Quoted 6 lines

Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 33 lines
Quoted 32 lines · Xuan

Hi,

Quoted 30 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 8 lines
Quoted 8 lines

time

Quoted 7 lines
Quoted 7 lines

ago

Quoted 6 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition.

Many

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines

examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and

split_header_size

Quoted 41 lines
Quoted 41 lines
Quoted 41 lines
Quoted 8 lines
Quoted 8 lines

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++
  1 file changed, 4 insertions(+)
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
    applications should be updated to use the ``dmadev``

library

instead,

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

with the underlying HW-functionality being provided by the
``ioat``

or

Quoted 21 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled

and

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines

protocol type configured, PMD will split the ingress packets into two separate regions.
"

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

"split_hdr_size".

Quoted 2 lines

But no single PMD actually supports
RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078

Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 1 line

-

2-w

Quoted 5 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

was

Quoted 10 lines
Quoted 6 lines
Quoted 1 line

configurable per-port.

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

Quoted 6 lines

Hi, thank you for Cc'ing

Yes, you are right, we have two splitting offloads, and these ones have the different intentions. As there are no PMDs actually handling RTE_ETH_RX_OFFLOAD_HEADER_SPLIT, there should be no objections for this deprecation.

Thanks for helping review.

For 'BUFFER_SPLIT', I think it is clear. As Ferruh explains, it is used for splitting packets into multi-segments and multi-mempools based on rte_eth_rxseg_split. Right now only mlx5 supports this offload.

For 'HEADER_SPLIT', IMO it is used for splitting packets into two segments based on the split_hdr_size in rte_eth_rxmode. And header split does not support split into multi-mempools(the same mempool).

So at this level, we can say that header split and buffer split are the same intention (split packets). The functions supported by header split have been covered by buffer split. And no PMD actually supports 'HEADER_SPLIT'.

Please corrects me if my understanding is wrong.

Thanks,
Xuan

Quoted 3 lines
#14 lore raw
Quoted 177 lines · Xuan

Hi,

Quoted 168 lines

-----Original Message-----
From: Slava Ovsiienko <[email protected]>
Sent: Friday, July 15, 2022 12:56 AM
To: Ferruh Yigit <[email protected]>; Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: RE: [PATCH] doc: announce header split deprecation

Quoted 5 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, July 14, 2022 18:58
To: Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected];

Slava

Quoted 6 lines

Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 33 lines
Quoted 32 lines · Xuan

Hi,

Quoted 30 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 8 lines
Quoted 8 lines

time

Quoted 7 lines
Quoted 7 lines

ago

Quoted 6 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition.

Many

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines

examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and

split_header_size

Quoted 41 lines
Quoted 41 lines
Quoted 41 lines
Quoted 8 lines
Quoted 8 lines

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
   doc/guides/rel_notes/deprecation.rst | 4 ++++
   1 file changed, 4 insertions(+)
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
     applications should be updated to use the ``dmadev``

library

instead,

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

with the underlying HW-functionality being provided by the
``ioat``

or

Quoted 21 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled

and

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines

protocol type configured, PMD will split the ingress packets into two separate regions.
"

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

"split_hdr_size".

Quoted 2 lines

But no single PMD actually supports
RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078

Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 1 line

-

2-w

Quoted 5 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

was

Quoted 10 lines
Quoted 6 lines
Quoted 1 line

configurable per-port.

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

Quoted 6 lines

Hi, thank you for Cc'ing

Yes, you are right, we have two splitting offloads, and these ones have the different intentions. As there are no PMDs actually handling RTE_ETH_RX_OFFLOAD_HEADER_SPLIT, there should be no objections for this deprecation.

Quoted 8 lines

Thanks for helping review.

For 'BUFFER_SPLIT', I think it is clear. As Ferruh explains, it is used for splitting packets into multi-segments and multi-mempools based on rte_eth_rxseg_split. Right now only mlx5 supports this offload.

For 'HEADER_SPLIT', IMO it is used for splitting packets into two segments based on the split_hdr_size in rte_eth_rxmode. And header split does not support split into multi-mempools(the same mempool).

I looks like we don't have much details on the intention of the 'HEADER_SPLIT', it is not well documented.

Quoted 13 lines

So at this level, we can say that header split and buffer split are the same intention (split packets). The functions supported by header split have been covered by buffer split. And no PMD actually supports 'HEADER_SPLIT'.

Please corrects me if my understanding is wrong.

Thanks,
Xuan

Quoted 3 lines
author
#15 lore raw
Quoted 200 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, July 15, 2022 7:34 PM
To: Ding, Xuan <[email protected]>; Slava Ovsiienko <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 137 lines · Xuan

Hi,

Quoted 135 lines

-----Original Message-----
From: Slava Ovsiienko <[email protected]>
Sent: Friday, July 15, 2022 12:56 AM
To: Ferruh Yigit <[email protected]>; Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: RE: [PATCH] doc: announce header split deprecation

Quoted 5 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Thursday, July 14, 2022 18:58
To: Ding, Xuan <[email protected]>; NBU-Contact-Thomas Monjalon (EXTERNAL) <[email protected]>; [email protected];

Slava

Quoted 6 lines

Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 33 lines
Quoted 32 lines · Xuan

Hi,

Quoted 30 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Thursday, July 14, 2022 9:25 PM
To: Ding, Xuan <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

14/07/2022 14:54, Ding, Xuan:

Quoted 9 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 6 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced

some

Quoted 8 lines
Quoted 8 lines

time

Quoted 7 lines
Quoted 7 lines

ago

Quoted 6 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition.

Many

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines

examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and

split_header_size

Quoted 41 lines
Quoted 41 lines
Quoted 41 lines
Quoted 8 lines
Quoted 8 lines

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
   doc/guides/rel_notes/deprecation.rst | 4 ++++
   1 file changed, 4 insertions(+)
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
     applications should be updated to use the ``dmadev``

library

instead,

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

with the underlying HW-functionality being provided by
the ``ioat``

or

Quoted 21 lines
Quoted 21 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst

Quoted 7 lines

``idxd`` dma drivers

doc/guides/rel_notes/deprecation.rst +6 -0
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled

and

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines

protocol type configured, PMD will split the ingress packets into two separate regions.
"

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

"split_hdr_size".

Quoted 2 lines

But no single PMD actually supports
RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

for this purpose.

Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 1 line

-

2-w

Quoted 3 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT

instead.

Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 2 lines

It is for multi-segments packet split. And it still needs a "proto_hdr" field in

rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 5 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 2 lines

So IMO the deprecation for header split is not relevant with buffer split. But

we can still clean the code.

Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and

it

Quoted 15 lines
Quoted 15 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT

was

Quoted 10 lines
Quoted 6 lines
Quoted 1 line

configurable per-port.

Thanks for your clarification. It's clearer now.
I was trying to figure out the whole history of header split, seems it is not enough.

Isn't the intention of 'RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT' & 'RTE_ETH_RX_OFFLOAD_HEADER_SPLIT' are different? Cc'ed Slava for more comment.

Hi, thank you for Cc'ing

Quoted 8 lines
Quoted 4 lines

Yes, you are right, we have two splitting offloads, and these ones have the different intentions. As there are no PMDs actually handling RTE_ETH_RX_OFFLOAD_HEADER_SPLIT, there should be no objections for this deprecation.

Thanks for helping review.

For 'BUFFER_SPLIT', I think it is clear. As Ferruh explains, it is used for

splitting packets into multi-segments and multi-mempools based on rte_eth_rxseg_split.

Quoted 2 lines

Right now only mlx5 supports this offload.

Quoted 8 lines
Quoted 1 line

For 'HEADER_SPLIT', IMO it is used for splitting packets into two segments

based on the split_hdr_size in rte_eth_rxmode.

Quoted 1 line

And header split does not support split into multi-mempools(the same

mempool).

Quoted 1 line

I looks like we don't have much details on the intention of the 'HEADER_SPLIT', it is not well documented.

Yes, since no PMD supports this offload, we should remove it to make code cleaner.

Thanks,
Xuan

Quoted 16 lines
Quoted 2 lines

So at this level, we can say that header split and buffer split are the same

intention (split packets).

Quoted 1 line

The functions supported by header split have been covered by buffer split.

And no PMD actually supports 'HEADER_SPLIT'.

Quoted 11 lines

Please corrects me if my understanding is wrong.

Thanks,
Xuan

Quoted 3 lines
#16 lore raw
In reply to Thomas Monjalon
Quoted 93 lines · Thomas Monjalon

14/07/2022 14:54, Ding, Xuan:

Quoted 76 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 55 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 24 lines
Quoted 24 lines
Quoted 13 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++
  1 file changed, 4 insertions(+)
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 9 lines
doc/guides/rel_notes/deprecation.rst +6 -0
@@ -125,3 +125,7 @@ Deprecation Notices
    applications should be updated to use the ``dmadev`` library instead,
    with the underlying HW-functionality being provided by the ``ioat`` or
    ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 6 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 1 line

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

Quoted 11 lines

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the
rte_eth_rxseg_split structure reserved field to specify header protocol type. With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using "split_hdr_size". But no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT for this purpose. So we finally decide to deprecate this flag.

Quoted 4 lines

http://patchwork.dpdk.org/project/dpdk/patch/[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field in rte_eth_rxmode to configure split location.

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split. But we can still clean the code. Hope it make things clearer.

Quoted 8 lines

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue,
while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

+1 tt is definitely a very good idea.

author
#17 lore raw

Hi,

Quoted 120 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Friday, July 15, 2022 8:14 PM
To: Thomas Monjalon <[email protected]>; Ding, Xuan <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; Zhang, Qi Z <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] doc: announce header split deprecation

Quoted 32 lines · Thomas Monjalon

14/07/2022 14:54, Ding, Xuan:

Quoted 31 lines

Hi,

From: Thomas Monjalon <[email protected]>

Quoted 28 lines

14/07/2022 07:50, Ding, Xuan:

Quoted 5 lines

From: Thomas Monjalon <[email protected]>

Quoted 4 lines

23/05/2022 16:20, [email protected]:

Quoted 3 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some

time

Quoted 5 lines
Quoted 5 lines

ago

Quoted 4 lines

to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable header split offload with the header size controlled using split_hdr_size in the same structure.

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but

some double-check that its value is 0.

Quoted 2 lines

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size

field

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

will be removed in DPDK 22.11.

---
  doc/guides/rel_notes/deprecation.rst | 4 ++++
  1 file changed, 4 insertions(+)
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 125

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +0 -0
index 4e5b23c53d..b8114f29ed 100644
@@ -125,3 +125,7 @@ Deprecation Notices
    applications should be updated to use the ``dmadev`` library

instead,

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

doc/guides/rel_notes/deprecation.rst Deprecation Notices · line 126

Quoted 1 line
doc/guides/rel_notes/deprecation.rst +0 -0
    with the underlying HW-functionality being provided by the ``ioat``

or

Quoted 33 lines
Quoted 33 lines
Quoted 32 lines
Quoted 9 lines
Quoted 9 lines

doc/guides/rel_notes/deprecation.rst Deprecation Notices

Quoted 7 lines
doc/guides/rel_notes/deprecation.rst +6 -0
    ``idxd`` dma drivers
+
+* ethdev: After bit-field header split was removed, the
+``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT``
+offload and the ``split_hdr_size`` field in structure
+``rte_eth_rxmode`` to enable header split offload are not
+supported in any

PMDs. They will be removed in DPDK 22.11.

Quoted 7 lines
Quoted 2 lines

It would have been good to talk about rte_eth_rxseg_split which is similar and configured per-queue.

Thanks for your suggestion.

But I'm a little confused, are you referring that I need to involve protocol

based buffer split?

Quoted 2 lines

About the deprecation of header split, I haven't realized its connection to

rte_eth_rxseg_split.

What???
In old versions of your patch "ethdev: introduce protocol type based header split"
you wrote:
"
A new proto field is introduced in the rte_eth_rxseg_split structure reserved field to specify header protocol type.
With Rx offload flag RTE_ETH_RX_OFFLOAD_HEADER_SPLIT enabled and protocol type configured, PMD will split the ingress packets into two separate regions.
"

Quoted 3 lines
Quoted 3 lines

It has a long history...
It was corrected in v4 that RTE_ETH_RX_OFFLOAD_HEADER_SPLIT is used to enable header split offload with the header size controlled using

"split_hdr_size".

Quoted 1 line
Quoted 1 line

But no single PMD actually supports

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT for this purpose.

Quoted 3 lines
Quoted 3 lines

So we finally decide to deprecate this flag.

http://patchwork.dpdk.org/project/dpdk/patch/20220402104109.472078-

2-

Quoted 4 lines
Quoted 4 lines

[email protected]/

In following series, I use RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT instead. It is for multi-segments packet split. And it still needs a "proto_hdr" field

in rte_eth_rxmode to configure split location.

Quoted 10 lines

I know this history because I was the one asking you to deprecate this. But it seems you didn't get the big picture.

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

Currently there are 2 acks, add more PMD maintainers to help review this deprecation notice for header split, thanks a lot!

I cannot say my feeling strong enough.

So IMO the deprecation for header split is not relevant with buffer split.

But we can still clean the code.

Quoted 8 lines
Quoted 1 line

Hope it make things clearer.

They are almost the same features.
So when deprecating one, it is important to mention what remains. If needed RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used and it is configured per-queue, while RTE_ETH_RX_OFFLOAD_HEADER_SPLIT was configurable per-port.

Quoted 4 lines
Quoted 2 lines

Andrew, Ferruh, do you agree to improve this deprecation notice by adding above information?

+1 tt is definitely a very good idea.

Thanks, v2 is on the way.

Regards,
Xuan

#18 lore raw

PATCH v2 1/1 doc: announce header split deprecation

In reply to [email protected]

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable per-port header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11. After DPDK 22.11 LTS, the RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used for per-queue Rx packet split offload, which is configured by rte_eth_rxseg_split.

---
v2:
* Refine doc to mention remained RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT.

---
 doc/guides/rel_notes/deprecation.rst | 7 +++++++
 1 file changed, 7 insertions(+)
doc/guides/rel_notes/deprecation.rst +7 -0
index 1e5e532e60..ae37559275 100644
@@ -158,3 +158,10 @@ Deprecation Notices
   applications should be updated to use the ``dmadev`` library instead,
   with the underlying HW-functionality being provided by the ``ioat`` or
   ``idxd`` dma drivers
+
+* ethdev: Since no single PMD supports ``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT`` offload
+and the ``split_hdr_size`` field in structure ``rte_eth_rxmode`` to enable per-port
+header split as they were designed, they will be removed in DPDK 22.11.
+After DPDK 22.11 LTS, the ``RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT`` can still be
+used for per-queue Rx packet split offload, and it is configured
+by ``rte_eth_rxseg_split``.
Signature · 1 line
2.17.1
Acked-by on v2 1/1 Applied
#19 lore raw

15/07/2022 22:30, [email protected]:

Quoted 17 lines

From: Xuan Ding <[email protected]>

RTE_ETH_RX_OFFLOAD_HEADER_SPLIT offload was introduced some time ago to substitute bit-field header_split in struct rte_eth_rxmode. It allows to enable per-port header split offload with the header size controlled using split_hdr_size in the same structure.

Right now, no single PMD actually supports RTE_ETH_RX_OFFLOAD_HEADER_SPLIT with above definition. Many examples and test apps initialize the field to 0 explicitly. The most of drivers simply ignore split_hdr_size since the offload is not advertised, but some double-check that its value is 0.

So the RTE_ETH_RX_OFFLOAD_HEADER_SPLIT and split_header_size field will be removed in DPDK 22.11. After DPDK 22.11 LTS, the RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT can still be used for per-queue Rx packet split offload, which is configured by rte_eth_rxseg_split.

Quoted 5 lines

This v2 is a lot better for the users to understand what happens. Fixed the indent and moved few words:

* ethdev: Since no single PMD supports ``RTE_ETH_RX_OFFLOAD_HEADER_SPLIT`` offload and the ``split_hdr_size`` field in structure ``rte_eth_rxmode`` to enable per-port header split, they will be removed in DPDK 22.11. The per-queue Rx packet split offload ``RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT`` can still be used, and it is configured by ``rte_eth_rxseg_split``.

Applied, thanks.