NVIDIA ConnectX-6 Adapter Cards Firmware Release Notes v20.34.1002
NVIDIA ConnectX-5 Adapter Cards Firmware Release Notes v16.35.3502 LTS

Known Issues

Ethernet Rate Limit per VF in RoCE Mode Limitations

Dual Port Device

Single Port Device

w/o LAG (TOTAL_VFS>32)

With LAG (TOTAL_VFS<32)

w/o LAG

w/o QoS

Full QoS

w/o QoS

Full QoS

w/o QoS

Full QoS

127

127

64

64

127

127

Ethernet Rate Limit per VF in InfiniBand Mode Limitations

Dual Port Device

Single Port Device

w/o LAG

w/o LAG

w/o QoS

Full QoS

w/o QoS

Full QoS

127

127

127

127

Known Issues

Internal Ref.

Issue

3200779

Description: Changing dynamic PCIe link width is not supported.

Workaround: N/A

Keywords: PCIe

Discovered in Version: 20.34.1002

2824427

Description: Running with a debug firmware reduces security as if token was applied.

Workaround: N/A

Keywords: Debug Firmware

Discovered in Version: 20.33.1048

3030570

Description: “crypto policy” access registry can be modified only by the INI file.

Workaround: N/A

Keywords: AES_XTS

Discovered in Version: 20.33.1048

2793880

Description: Checksum is not calculated correctly in IPoIP packet with LSO.

Workaround: N/A

Keywords: IPoIP, LSO, checksum

Discovered in Version: 20.33.1048

2864238

Description: VPD cannot be accessed after firmware upgrade or reset when the following sequence is performed:

  1. Upgrade to a new firmware and perform a cold reboot

  2. Downgrade to an old firmware

  3. Run fwreset

  4. Upgrade to a new firmware

  5. Run fwreset

Workaround: Run the upgrade or reset sequence as follow:

  1. Upgrade to a new firmware and perform a cold reboot

  2. Downgrade to an old firmware

  3. Run fwreset

  4. Upgrade to a new firmware

  5. Perform a cold reboot

Keywords: VDP

Discovered in Version: 20.32.1010

2616755

Description: Forward action for IPoIB is not supported on RX RDMA Flow Table.

Workaround: N/A

Keywords: Steering, IPoIB

Discovered in Version: 20.32.1010

2582094

Description: When performing a stress toggle test vs. IXIA, the IXIA side is not ready for few seconds.

Workaround: Wait for 1 sec between running the down and up commands.

Keywords: Auto-negotiation.

Discovered in Version: 20.31.1014

2446583

Description: On rare occasions, when both network devices are NVIDIA, PAM4 link will raise with several effective errors. These errors will not affect traffic once the link is up.

Workaround: Clear counters once the link is up

Keywords: Effective errors

Discovered in Version: 20.29.2002

SF 933911

Description: PXE boot will not function if the adapter card is connected to a NVIDIA Quantum™ based switch over an HDR active optical cable.

Workaround: Set KEEP_LINK_UP_ON_BOOT configuration to enable via mlxconfig.
For further information please contact Support.

Keywords: IB, Link Speed , Link Down, SDR , optical cable

Discovered in Version: 20.30.1004

2378593

Description: Sub 1sec firmware update (fast reset flow) is not supported when updating from previous releases to the current one. Doing so may cause network disconnection events.

Workaround: Use full reset flow for firmware upgrade/downgrade.

Keywords: Sub 1sec firmware update

Discovered in Version: 20.29.1016

2213356

Description: The following are the Steering Dump limitations:

  • Supported only on ConnectX-5 adapter cards

  • Requires passing the version (FW/Stelib/MFT) and device type to stelib

  • Re-format is not supported

  • Advanced multi-port feature is not supported – LAG/ROCE_AFFILIATION/MPFS_LB/ESW_LB (only traffic vhca <-> wire)

  • Packet types supported:

    • Layer 2 Eth

    • Layer 3 IPv4/Ipv6/Grh

    • Layer 4 TCP/UDP/Bth/GreV0/GreV1

    • Tunneling VXLAN/Geneve/GREv0/Mpls

  • FlexParser protocols are not supported (e.g AliVxlan/VxlanGpe etc..).

  • Compiles only on x86

Workaround: N/A

Keywords: Steering Bump

Discovered in Version: 20.29.1016

2365322

Description: When configuring adapter card's Level Scheduling, a QoS tree leaf (QUEUE_GROUP) configured with default rate_limit and default bw_share, may not obey the QoS restrictions imposed by any of the leaf’s ancestors.

Workaround: To prevent such a case, configure at least one of the following QoS attributes of a leaf: max_average_bw or bw_share

Keywords: QoS

Discovered in Version: 20.29.1016

2239632

Description: EDR linkup time might take up to 50sec when using HDR optical cable.

Workaround: N/A

Keywords: Cables, EDR

Discovered in Version: 20.28.1002

2145881

Description: FDR link is unstable when using an FDR cable in ports: #27-#34.

Workaround: N/A

Keywords: FDR, cables

Discovered in Version: 20.27.6008

2149437

Description: When the SLTP configuration is wrongly set, the “Bad status” explanation will not be presented (only error indication) to the user.

Workaround: N/A

Keywords: SLTP configuration

Discovered in Version: 20.27.6008

2071210

Description: mlxconfig query for the BOOT_INTERRUPT_DIS TLV shows a wrong value in the "current value" field.

Workaround: Use "next boot" indication to see the right value.

Keywords: mlxconfig

Discovered in Version: 20.27.1016

1796936

Description: 200GbE Optical cables in Auto-Negotiation mode work only in 200GbE speed.

Workaround: N/A

Keywords: Cables

Discovered in Version: 20.27.1016

1959529

Description: When HDR Active Copper cables are used between Quantum switches, or between Quantum switch and ConnectX-6 HCA, the counter indicating ‘Link Down’ may have a value other than zero, after the first time the cable is connected. As this may happened only at the first time, it is recommend to clear the counters after the cluster is brought up.

Workaround: Toggle the Active Copper or Optics cables as the switch performs a reset.

Keywords: Cables, BER

Discovered in Version: 20.27.1016

1959529

Description: Occasionally (up to 15% of connections), the link will go down when using ACC cables P/N: MCA1J00-H003E, MCA1J00-H004E and when connecting a Quantum switch to a Quantum switch.

Workaround: N/A

Keywords: Cables

Discovered in Version: 20.27.1016

1997329

Description: Downgrading from firmware v20.26.4012 to firmware v20.26.1040 and lower is not supported on Windows OSes using the mlxfwmanager tool.

Workaround: N/A

Keywords: mlxfwmanager, firmware downgrade

Discovered in Version: 20.26.4012

1930619

Description: PF_BAR2 and ATS cannot be enabled together, i.e. when PF_BAR2 is enabled, ATS cannot be enabled too.

Workaround: N/A

Keywords: ATS, SF, BAR2, Multi GVMI

Discovered in Version: 20.26.1040

-

Description: In rare cases, following a server powerup, a fatal error (device's health compromised) message might appear with ext_synd 0x8d1d. The error will be accompanied by a failure to use mlxconfig and in some cases flash burning tools.

Workaround: N/A

Keywords: mlxconfig, flash tool, ext_synd 0x8d1d

Discovered in Version: 20.26.1040

1796936

Description: HDR split cables support only HDR speed.

Workaround: N/A

Keywords: Link Speed, cables, Break-Out cables

Discovered in Version: 20.26.1040

1750460 / 2063991

Description: BER issues might occur when using ConnectX-6 adapter cards in 100GbE link speed, and connecting with and 3rd party switch systems.

Workaround: N/A

Keywords: BER, 100GbE, Spectrum-2

Discovered in Version: 20.26.1040

1918749

Description: mlxlink tool displays a wrong speed when using ETH cables on ConnectX-6 adapter cards.

Workaround: N/A

Keywords: mlxlink

Discovered in Version: 20.26.1040

1901198

Description: Firmware is not loaded on Multi-Host setups after reboot.

Workaround: N/A

Keywords: Firmware load, Multi-Host

Discovered in Version: 20.26.1040

1842278

Description: DC LAG can function only in case there is a single PF per port without any active VFs.

Workaround: N/A

Keywords: DC LAG

Discovered in Version: 20.26.1040

1796628

Description: Due to performance considerations, unicast loopback traffic will go through the NIC SX tables, and multicast loopback traffic will skip the NIC SX tables.

Workaround: N/A

Keywords: Performance, unicast loopback traffic, multicast loopback traffic

Discovered in Version: 20.26.1040

1797493

Description: Firmware asserts may occur when setting the PF_BAR2_SIZE value higher than the maximum supported size.

Workaround: Configure within limits (NIC PF_BAR_SIZE <= 4).

Keywords: Multi-GVMI, Sub-Function, SFs, BAR2

Discovered in Version: 20.26.1040

-

Description: Coherent Accelerator Processor Interface (CAPI) in ConnectX-6 firmware v20.25.7020 and above has low test coverage, however, it has no known issues.

Workaround: N/A

Keywords: CAPI

Discovered in Version: 20.25.7020

-

Description: HDR optical cables and Split cables support only HDR speed.

Workaround: N/A

Keywords: Link Speed, cables, Break-Out cables

Discovered in Version: 20.25.6000

1755286

Description: Port speed may change to SDR spontaneously, without a clear reason.

Workaround: Keep the "keep_ib_link_up" bit at 0 in NVconfig to make sure the port is raised with the correct speed.

Keywords: SDR, port speed

Discovered in Version: 20.25.2006

1774135

Description: PXE boot is not functional when connecting a splitter cable to the host.

Workaround: Update the SM as follow:

  • MLNX_OFED SM:

    • Set the default partition in the SM partitions.conf file as shown in the example below:

Default=0x7fff,ipoib,rate=5:ALL=full;

Note: "rate" must be set to "5" regardless to the other flags values.

  • MLNX-OS SM:

Run the following CLI commands:

no ib sm

ib partition Default rate 5 ib sm

  • UFM SM:

Use REST API to change default partition rate:

PUT https://<some IP>/ufmRest/resources/networks/management

{

"qos_parameters": { "rate_limit": 900

}

}

As a result, /opt/ufm/files/conf/opensm/partitions.conf will include the following line:

management=0x7fff,ipoib, sl=0,rate=5, defmember=full : ALL, ALL_SWITCHES=full,SELF=full;

Keywords: PXE boot, splitter cable

Discovered in Version: 20.25.2006

1768814/1772474

Description: Due to hardware limitation, REG_C cannot be passed over loopback when the FDB action is forwarded to multiple destinations.

Workaround: N/A

Keywords: Connection-Tracking

Discovered in Version: 20.25.2006

1770736

Description: When a PF or ECPF with many VFs (SR-IOV), and/or SFs (Multi-GVMI) triggers an FLR, PCIe completion timeout might occur.

Workaround: Increase the PCIe completion timeout.

Keywords: Multi-GVMI, SR-IOV, Sub-Function, Virtual Function, PF FLR

Discovered in Version: 20.25.2006

1716334

Description: When mlxconfig.PF_BAR2_EN is enabled, configuring more than 255 PCI functions will raise an assert.

Workaround: When working with BAR2, configure SR-IOV to align to the 255 PCI functions limitation.

mlxconfig.NUM_OF_VFS controls the number of configured SR-IOV VFs. e.g.:

  • Smart NICs: 2 External Host PFs, 2 ARM ECPFs, 125 VFs per PF.

  • Non-smart NICs: 2 External Host PFs, 126 VFs per PF

Keywords: Multi-GVMI, PF_BAR2_EN, Sub-Functions, SR-IOV, VFs

Discovered in Version: 20.25.1500 [Beta]

1699214

Description: NODNIC VF is partially tested. It is fully tested only in ConnectX-5 adapter cards.

Workaround: N/A

Keywords: NODNIC VF

Discovered in Version: 20.25.1500 [Beta]

1699214

Description: NODNIC VF is partially tested. It is fully tested only in ConnectX-5 adapter cards.

Workaround: N/A

Keywords: NODNIC VF

Discovered in Version: 20.25.1500 [Beta]

-

Description: The supported length of HDR copper cables is currently up to 2M.

Workaround: N/A

Keywords: HDR cables

Discovered in Version: 20.25.1500 [Beta]

-

Description: In Ethernet mode, at 10/40GbE speeds, only NO-FEC in Force mode is supported. Other user configurations are overridden.

Workaround: N/A

Keywords: Ethernet, 10GbE, 40GbE, RS-FEC

Discovered in Version: 20.25.1500 [Beta]

1574876

Description: DC RoCE LAG is functional only if the router posts VRRP address as the source MAC.

Workaround: N/A

Keywords: DC RoCE LAG

Discovered in Version: 20.25.1500 [Beta]

1498399

Description: If the XRC switches between SRQ/RMPs while there is an outstanding ODP on the responder XRC QP, a CQE with an error might be generated (that is not a PFAULT abort).

Workaround: N/A

Keywords: XRC SRQ/RMP ODP

Discovered in Version: 20.25.1500 [Beta]

-

Description: In some cases, the power consumption might be 10% higher than what is stated in the adapter cards User Manual.

Workaround: Power consumption will be aligned with the User Manual statement in the next release

Keywords: Power consumption

Discovered in Version: 20.25.1500 [Beta]

1546492

Description: Executing the update_lid command while the IB port sniffer utility is active can stop the utility.

Workaround: N/A

Keywords: IB Sniffer

Discovered in Version: 20.25.1500 [Beta]

1537898

Description: Initializing a function while the IB port sniffer utility is active can stop the utility.

Workaround: N/A

Keywords: IB Sniffer

Discovered in Version: 20.25.1500 [Beta]

1414290

Description: When getting an inline scatter CQE on IB striding RQ, the stride index in the CQE will be zero.

Workaround: N/A

Keywords: Scatter CQE

Discovered in Version: 20.25.1500 [Beta]

1332714/1345824

Description: The maximum “read” size of MTRC_STDB is limited to 272 Bytes.

Workaround: Set the MTRC_STDB.read_size to the maximum value of 0x110=272 Bytes

Keywords: Access register, MTRC_STDB, tracer to dmesg, fwtrace to dmesg

Discovered in Version: 20.25.1500 [Beta]

1408994

Description: FTE with both forward (FWD) and encapsulation (ENCAP) actions is not supported in the SX NIC Flow Table.

Workaround: N/A

Keywords: SX NIC Flow Table

Discovered in Version: 20.25.1500 [Beta]

1027553

Description: While using e-switch vport sVLAN stripping, the RX steering values on the sVLAN might not be accurate.

Workaround: N/A

Keywords: e-sw vport sVLAN stripping, RX steering

Discovered in Version: 20.25.1500 [Beta]

1799917

Description: Untagged CVLAN packets in the Steering Flow Tables do not match the SVLAN tagged packets.

Workaround: N/A

Keywords: Steering Flow Tables, CVLAN/SVLAN packets

Discovered in Version: .20.25.1500 [Beta]

1277762

Description: An Ethernet multicast loopback packet is not counted (even if it is not a local loopback packet) when running the nic_receive_steering_discard command.

Workaround: N/A

Keywords: Ethernet multicast loopback packet

Discovered in Version: 20.25.1500 [Beta]

1306342

Description: Signature-accessing WQEs sent locally to the NVMeF target QPs that encounter signature errors, will not send a SIGERR CQE.

Workaround: N/A

Keywords: Signature-accessing WQEs, NVMeF target

Discovered in Version: 20.25.1500 [Beta]

1168594

Description: RoCE Dual Port Mode (a.k.a Multi-Port vHCA: MPV) is not supported in Multi-Host setups.

Workaround: N/A

Keywords: Multi-Port vHCA, Multi-Host

Discovered in Version: 20.25.1500 [Beta]

1072337

Description: If a packet is modified in e-sw flow steering, the SX sniffer Flow Table (of the VF) will see the sniffed packet after the modification.

Workaround: N/A

Keywords: SX sniffer Flow Table

Discovered in Version: 20.25.1500 [Beta]

1171013

Description: Signature Handover Operations is not supported when FPP (Function-Per-Port) mode is disabled.

Workaround: N/A

Keywords: Signature Handover Operations, FPP

Discovered in Version: 20.25.1500 [Beta]

1059975

Description: NVMeF limitation:

  • Transaction size - up to 128KB per IO (non-inline)

  • Support up to 16K connections

  • Support single namespace per drive

  • Staging buffer size must be at least 16MB in order to allow SRQ size of 64 entries

Workaround: N/A

Keywords: NVMeF

Discovered in Version: 20.25.1500 [Beta]

© Copyright 2023, NVIDIA. Last updated on May 23, 2023.