Skip to contentSkip to conversation

virtio-net: correct conditions for devices to validate packet checksum

PATCH In review Heng Qi 4 participants 10 messages Jan 3, 2024 – Feb 27, 2024

lore
Share & tools
lore

Heng Qi

author
#0 lore raw

PATCH v8 1/1 virtio-net: correct conditions for devices to validate packet checksum

There is a historical error in virtio spec:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

 device-types/net/description.tex | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +3 -11
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
   \field{num_buffers} is one, then the entire packet will be
   contained within this buffer, immediately following the struct
   virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the
-  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
+\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
   set: if so, device has validated the packet checksum.
   In case of multiple encapsulated protocols, one level of checksums
   has been validated.
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
 packet together, such that at least \field{num_buffers} are
 observed by driver as used.
 
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.
 
 If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set
 \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
 not less than the length of the headers, including the transport
 header.
 
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
 \drivernormative{\paragraph}{Processing of Incoming
 Packets}{Device Types / Network Device / Device Operation /
 Processing of Incoming Packets}
Signature · 18 lines
1.8.3.1


This publicly archived list offers a means to provide input to the
OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and
to minimize spam in the list archive, subscription is required
before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/
author
#1 lore raw
Quoted 37 lines · Heng Qi

There is a historical error in virtio spec:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Quoted 6 lines

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Hi Michael, Cornelia.

Could you please open a voting for this fix?

Thanks a lot!

Quoted 63 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

  device-types/net/description.tex | 14 +++-----------
  1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +3 -5
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
    \field{num_buffers} is one, then the entire packet will be
    contained within this buffer, immediately following the struct
    virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the
-  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
+\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
    set: if so, device has validated the packet checksum.
    In case of multiple encapsulated protocols, one level of checksums
    has been validated.
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  packet together, such that at least \field{num_buffers} are
  observed by driver as used.
 
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.
 
  If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set
  \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  not less than the length of the headers, including the transport
  header.

device-types/net/description.tex \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network · line 845

Quoted 10 lines
device-types/net/description.tex +0 -6
 
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
  \drivernormative{\paragraph}{Processing of Incoming
  Packets}{Device Types / Network Device / Device Operation /
  Processing of Incoming Packets}

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

Reviewed-by
#2 lore raw
Quoted 46 lines · Heng Qi
Quoted 42 lines · Heng Qi

There is a historical error in virtio spec:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Quoted 7 lines
Quoted 1 line

Hi Michael, Cornelia.

Could you please open a voting for this fix?

Thanks a lot!

Patch looks good to me.

Thanks

Quoted 83 lines
Quoted 64 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

  device-types/net/description.tex | 14 +++-----------
  1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +3 -5
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
    \field{num_buffers} is one, then the entire packet will be
    contained within this buffer, immediately following the struct
    virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the
-  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
+\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
    set: if so, device has validated the packet checksum.
    In case of multiple encapsulated protocols, one level of checksums
    has been validated.
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  packet together, such that at least \field{num_buffers} are
  observed by driver as used.
 
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.
 
  If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set
  \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  not less than the length of the headers, including the transport
  header.

device-types/net/description.tex \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network · line 845

Quoted 10 lines
device-types/net/description.tex +0 -6
 
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
  \drivernormative{\paragraph}{Processing of Incoming
  Packets}{Device Types / Network Device / Device Operation /
  Processing of Incoming Packets}

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Quoted 9 lines

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

author
#3 lore raw
In reply to Heng Qi
Quoted 45 lines · Heng Qi
Quoted 42 lines · Heng Qi

There is a historical error in virtio spec:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Quoted 5 lines
Quoted 1 line

Hi Michael, Cornelia.

Could you please open a voting for this fix?

Hi Michael and Cornelia.

Sincerely asking if you missed this message?

In addition, Jason recently responded with his reviewed-by tag. Do I need to publish a new version carrying his tag (thanks Jason)?

Thanks.

Quoted 75 lines

Thanks a lot!

Quoted 72 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

  device-types/net/description.tex | 14 +++-----------

  1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +0 -0
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network \field{num_buffers} is one, then the entire packet will be contained within this buffer, immediately following the struct virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the -  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be +\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
    set: if so, device has validated the packet checksum. In case of multiple encapsulated protocols, one level of checksums has been validated.

device-types/net/description.tex +0 -0
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network
  packet together, such that at least \field{num_buffers} are observed by driver as used.
  -If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set -\field{flags} to zero and SHOULD supply a fully checksummed

device-types/net/description.tex +2 -1
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.

    If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.

device-types/net/description.tex +0 -0
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network
  not less than the length of the headers, including the transport header.
  -If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the

device-types/net/description.tex +0 -5
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
Quoted 8 lines
Quoted 3 lines

  \drivernormative{\paragraph}{Processing of Incoming Packets}{Device Types / Network Device / Device Operation / Processing of Incoming Packets}

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

#4 lore raw
Quoted 67 lines · Heng Qi
Quoted 66 lines · Heng Qi
Quoted 62 lines · Heng Qi

There is a historical error in virtio spec:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates
the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially
checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped
to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets.
When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that
packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will
still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually
expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had
nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below
desctiption was added incorrectly:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be
controlled by whether GUEST_CSUM is negotiated, but only controlled by whether
rx checksum offload is enabled on the OS side. The state of this rx checksum
offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be
added to dev->hw_features. When the user turns off rx checksum offload through
ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is,
all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1]
https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Quoted 7 lines
Quoted 3 lines

Hi Michael, Cornelia.

Could you please open a voting for this fix?

Hi Michael and Cornelia.

Sincerely asking if you missed this message?

I've been back from end-of-year PTO only this week...

I'm currently waiting for a confirmed updated schedule for the infrastructure migration, as I don't want to risk having the platform out of order in the middle of the voting period.

Quoted 3 lines

In addition, Jason recently responded with his reviewed-by tag. Do I need to publish a new version carrying his tag (thanks Jason)?

No, we can add his R-b when pushing to git.

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

author
#5 lore raw
Quoted 70 lines · Cornelia Huck
Quoted 68 lines · Heng Qi
Quoted 67 lines · Heng Qi
Quoted 55 lines · Heng Qi

There is a historical error in virtio spec:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST*
set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates
the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially
checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped
to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets.
When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that
packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will
still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually
expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had
nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below
desctiption was added incorrectly:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST*
set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be
controlled by whether GUEST_CSUM is negotiated, but only controlled by whether
rx checksum offload is enabled on the OS side. The state of this rx checksum
offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be
added to dev->hw_features. When the user turns off rx checksum offload through
ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is,
all packets will be marked as CHECKSUM_NONE.

Quoted 7 lines

Related Link
=============
[1]
https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Hi Michael, Cornelia.

Could you please open a voting for this fix?

Quoted 8 lines
Quoted 3 lines

Hi Michael and Cornelia.

Sincerely asking if you missed this message?

I've been back from end-of-year PTO only this week...

I'm currently waiting for a confirmed updated schedule for the infrastructure migration, as I don't want to risk having the platform out of order in the middle of the voting period.

I got it. It depends on when the infrastructure is migrated, thanks for the sync.

Quoted 4 lines
Quoted 3 lines

In addition, Jason recently responded with his reviewed-by tag. Do I need to publish a new version carrying his tag (thanks Jason)?

No, we can add his R-b when pushing to git.

Ok.

Thanks.

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

Reviewed-by on v8 1/1
#6 lore raw
In reply to Heng Qi
Quoted 42 lines · Heng Qi

There is a historical error in virtio spec:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Quoted 4 lines
Quoted 71 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

 device-types/net/description.tex | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +3 -12
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
   \field{num_buffers} is one, then the entire packet will be
   contained within this buffer, immediately following the struct
   virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the
-  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
+\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
   set: if so, device has validated the packet checksum.
   In case of multiple encapsulated protocols, one level of checksums
   has been validated.
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
 packet together, such that at least \field{num_buffers} are
 observed by driver as used.
 
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.
 
 If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set
 \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
 not less than the length of the headers, including the transport
 header.
 
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
 \drivernormative{\paragraph}{Processing of Incoming
 Packets}{Device Types / Network Device / Device Operation /
 Processing of Incoming Packets}
--
Quoted 2 lines

1.8.3.1

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

author
#7 lore raw
Quoted 38 lines · Xuan Zhuo
Quoted 37 lines · Heng Qi

There is a historical error in virtio spec:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
"If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Quoted 6 lines
Quoted 6 lines

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Hi Cornelia,

Can this proposal be opened for voting?

Thank you very much!

Quoted 85 lines
Quoted 3 lines
Quoted 71 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

  device-types/net/description.tex | 14 +++-----------
  1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +3 -12
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
    \field{num_buffers} is one, then the entire packet will be
    contained within this buffer, immediately following the struct
    virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the
-  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
+\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
    set: if so, device has validated the packet checksum.
    In case of multiple encapsulated protocols, one level of checksums
    has been validated.
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  packet together, such that at least \field{num_buffers} are
  observed by driver as used.
 
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.
 
  If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set
  \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
  not less than the length of the headers, including the transport
  header.
 
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
  \drivernormative{\paragraph}{Processing of Incoming
  Packets}{Device Types / Network Device / Device Operation /
  Processing of Incoming Packets}
--
Quoted 2 lines

1.8.3.1

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Quoted 8 lines

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]

author
#8 lore raw
Quoted 46 lines · Heng Qi
Quoted 43 lines · Xuan Zhuo
Quoted 42 lines · Heng Qi

There is a historical error in virtio spec:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets. When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below desctiption was added incorrectly:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be controlled by whether GUEST_CSUM is negotiated, but only controlled by whether rx checksum offload is enabled on the OS side. The state of this rx checksum offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be added to dev->hw_features. When the user turns off rx checksum offload through ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is, all packets will be marked as CHECKSUM_NONE.

Related Link
=============
[1] https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Quoted 12 lines
Quoted 1 line
Quoted 1 line

Hi Cornelia,

Can this proposal be opened for voting?

Thank you very much!

Quoted 4 lines
Quoted 3 lines

Hi Cornelia.

Did you miss this email? :)

Both Jason and Xuan have replied to Reviewed-by tag. Can we initiate a vote?

Thanks!

Quoted 103 lines
Quoted 86 lines
Quoted 73 lines

---
v7->v8:
- Drop the new FULLY_CSUM feature and optimize the description. @Jason

v6->v7:
- FULLY_CSUM no longer depends on GUEST_CSUM.
- The description related to DATA_VALID has nothing to do with whether GUEST_CSUM is negotiated or not. @Jason

v5->v6:
- Rewrite and clarify patch description. @Michael

v4->v5:
- Remove the modification to the GUEST_CSUM. @Jason
- The description of this feature has been reorganized for greater clarity. @Michael

v3->v4:
- Streamline some repetitive descriptions. @Jason
- Add how features should work, when to be enabled, and overhead. @Jason @Michael

v2->v3:
- Add a section named "Driver Handles Fully Checksummed Packets" and more descriptions. @Michael

v1->v2:
- Modify full checksum functionality as a configurable offload that is initially turned off. @Jason

  device-types/net/description.tex | 14 +++-----------

  1 file changed, 3 insertions(+), 11 deletions(-)
device-types/net/description.tex +0 -0
index aff5e08..14978c0 100644
@@ -723,8 +723,7 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network \field{num_buffers} is one, then the entire packet will be contained within this buffer, immediately following the struct virtio_net_hdr.
-\item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the -  VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be +\item The VIRTIO_NET_HDR_F_DATA_VALID bit in \field{flags} can be
    set: if so, device has validated the packet checksum. In case of multiple encapsulated protocols, one level of checksums has been validated.

device-types/net/description.tex +0 -0
@@ -783,9 +782,8 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network
  packet together, such that at least \field{num_buffers} are observed by driver as used.

device-types/net/description.tex +2 -3
-If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device MUST set
-\field{flags} to zero and SHOULD supply a fully checksummed
-packet to the driver.
+If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device SHOULD supply
+a fully checksummed packet to the driver.

  If VIRTIO_NET_F_GUEST_TSO4 is not negotiated, the device MUST NOT set \field{gso_type} to VIRTIO_NET_HDR_GSO_TCPV4.

device-types/net/description.tex +0 -0
@@ -842,12 +840,6 @@ \subsubsection{Processing of Incoming 

Packets}\label{sec:Device Types / Network
  not less than the length of the headers, including the transport header.

device-types/net/description.tex +0 -6
-If the VIRTIO_NET_F_GUEST_CSUM feature has been negotiated, the
-device MAY set the VIRTIO_NET_HDR_F_DATA_VALID bit in
-\field{flags}, if so, the device MUST validate the packet
-checksum (in case of multiple encapsulated protocols, one level
-of checksums is validated).
-
Quoted 6 lines

  \drivernormative{\paragraph}{Processing of Incoming Packets}{Device Types / Network Device / Device Operation / Processing of Incoming Packets}

Signature · 1 line
1.8.3.1

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Quoted 8 lines

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Quoted 8 lines

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/

#9 lore raw
Quoted 80 lines · Heng Qi
Quoted 79 lines · Heng Qi
Quoted 59 lines · Xuan Zhuo
Quoted 57 lines · Heng Qi

There is a historical error in virtio spec:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

Currently in Linux and virtio-related implementations, the device validates
the packet checksum and sets DATA_VALID regardless of whether VIRTIO_NET_F_GUEST_CSUM is negotiated.

Please refer to the following summary and thread[1] for details and reasons:

Summary
=============
1. GUEST_CSUM at virtio spec 0.95 is intended to be compatible with partially
checksummed packets (NEEDS_CSUM <-> CHECKSUM_PARTIAL). So GUEST_CSUM is mapped
to NETIF_F_RXCSUM.
GUEST_CSUM only indicates whether the driver handles partially checksummed packets.
When XDP is loaded, the GUEST_CSUM's offload will be disabled, which means that
packets have NEEDS_CSUM set will be dropped, and packets have DATA_VALID set will
still be received.

2. When DATA_VALID was added to Linux in 2011[2] and virtio1.0, it was actually
expected that rx checksum offload (the driver can set CHECKSUM_UNNECESSARY) had
nothing to do with whether GUEST_CSUM was negotiated. But due to an error, below
desctiption was added incorrectly:
   "If VIRTIO_NET_F_GUEST_CSUM is not negotiated, the device *MUST* set flags
    to zero and SHOULD supply a fully checksummed packet to the driver."

3. We now hope to correct this error. Let the setting of DATA_VALID not be
controlled by whether GUEST_CSUM is negotiated, but only controlled by whether
rx checksum offload is enabled on the OS side. The state of this rx checksum
offload is not aware of the device.

4. [Optional] NETIF_F_RXCSUM corresponding to rx checksum offload should be
added to dev->hw_features. When the user turns off rx checksum offload through
ethtool -K, neither NEEDS_CSUM nor DATA_VALID should be taken care of, that is,
all packets will be marked as CHECKSUM_NONE.

Quoted 7 lines
Quoted 7 lines

Related Link
=============
[1]
https://lists.oasis-open.org/archives/virtio-comment/202312/msg00135.html
[2] 10a8d94a9574 ("virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID")

Hi Cornelia,

Can this proposal be opened for voting?

Thank you very much!

Quoted 3 lines
Quoted 3 lines
Quoted 8 lines
Quoted 1 line
Quoted 1 line

Hi Cornelia.

Did you miss this email? :)

Both Jason and Xuan have replied to Reviewed-by tag. Can we initiate a vote?

We currently can't do any voting (OASIS infrastructure update), and we probably have to sort out the fallout from the new platform once it's up.

This publicly archived list offers a means to provide input to the OASIS Virtual I/O Device (VIRTIO) TC.

In order to verify user consent to the Feedback License terms and to minimize spam in the list archive, subscription is required before posting.

Subscribe: [email protected]
Unsubscribe: [email protected]
List help: [email protected]
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/