Known Issues

127 VF per PF (254 functions)

127

• 127 VF per PF (254 functions)

• 512 PF+VF+SF per PF (1024 functions)

• 127 VF (127 functions)

• 512 PF+VF+SF per PF (512 functions)

4884739

Description: Link failures may occasionally be observed at PAM4 speeds over optical interfaces in rare cases.

Workaround: Toggle the port (admin down/up).

Keywords: PAM4 speeds, optical interfaces

Detected in version:

40.47.1026

4604969

Description: Probe packets might be dropped at the transmission stage when multiple congestion control flows are active.

Workaround: N/A

Keywords: PCC, RTT, probe

Detected in version:

40.47.1026

4496642

Description: The timestamps (t2, t4) of the received RTT probes are taken from the free-running clock, even when ROCE_CC_RTT_TIMESTAMP_FORMAT is set to 0x02.
The format of all RTT probe timestamps can be found in HCA_CAP.rtt_timestamp_format.

Workaround: N/A

Keywords: RTT RTC timestamp

Detected in version:

40.47.1026

4705241

Description: When Quantum-2 is part of an XDR topology, serving as a leaf switch connected to NDR-based hosts, a bandwidth degradation of approximately 3–7 Gb/s is expected.

Workaround: N/A

Keywords: XDR, NDR, Quantum-2

Detected in version:

40.47.1026

4705948

Description: When using DC as the InfiniBand transport type to perform an ib_read RDMA operation between ConnectX-7 (NDR) and ConnectX-8, a bandwidth degradation of approximately 25% may be observed when using a low number of QPs (1–16). The performance degradation diminishes as the number of QPs increases.

Workaround: N/A

Keywords: DC, ib_read RDMA, NDR, performance

Detected in version:

40.47.1026

4685736

Description: Creating a DPA process that allocates a 128 MB data segment and loads a dynamic library may fail with syndrome 0xdc30ac.

Workaround: Limit the DPA application’s data segment size to 64 MB.

Keywords: DPA

Detected in version:

40.47.1026

4618452

Description: When only a partial number of fiber lanes are connected to a module (i.e., not all supported lanes are populated), the module re-insertion process may experience an extended link-up time of approximately one minute. This occurs because the module requires additional time to detect that only a subset of the lanes is connected.

Workaround: N/A

Keywords: Cables

Detected in version:

40.47.1026

4270913

Description: For DC InfiniBand transport, using SRQs requires distributing QPs across four SRQs. Each QP must be assigned to its own SRQ when there are up to four QPs.

Workaround: If more than four QPs are used, they must be evenly distributed across the four available SRQs.

Keywords: DC InfiniBand transport, SRQ, QPs

Detected in version:

40.47.1026

4665802

Description: Due to a re-burst scheduler limitation, the expected “speed of light” value for single QP RDMA Write on ConnectX-8 devices cannot be calculated. Currently, verification is limited to identifying performance degradations relative to previous firmware versions.

Workaround: N/A

Keywords: Single QP RDMA Write

Detected in version:

40.47.1026

4270913

Description: When using DC InfiniBand transport, SRQ usage requires distributing QPs across four SRQs. This means assigning each QP to its own SRQ when there are up to four QPs.

Workaround: If there are more than four QPs, they must be distributed equally among the four available SRQs.

Keywords: DC InfiniBand transport, SRQ, QPs

Detected in version:

40.45.1200

4412310

Description: Split operation (port splitting) on the second port is not supported on the ConnectX-8 SuperNIC model 900-9X81Q-00CN-ST0.

Workaround: N/A

Keywords: Port splitting

Detected in version:

40.45.1020

4394475

Description: The existing congestion control configuration applies globally, rather than on a per-priority basis.

Workaround: Ensure that the configuration values for all priorities are aligned in either mlxconfig ROCE_CC_PRIO_MASK_P$port or sysfs ecn/roce_rp/enable/$port.

Keywords: Congestion control, ROCE_CC_PRIO

Detected in version:

40.45.1020

4240828

Description: Simultaneous access to the same PDDR module info page is not supported.

Workaround: N/A

Keywords: PDDR module

Detected in version:

40.45.1020

4318063

Description: When running the PTP4L application, the path delay displays inconsistent values after each fwreset and rerun, resulting in a non-constant PPS offset that fails to meet Class B/C requirements.

Workaround: Manually perform a port link reset after fwreset.

Keywords: PTP

Detected in version:

40.45.1020

4163634

Description: When connecting a Quantum-3 switch system (with a port split into 8 ports) to a ConnectX-8 single port SuperNIC, the link will not be established.

Workaround: Configure the Quantum-3 switch system port to be split into 2 or 4 ports, or set the ConnectX-8 to operate in multiplane mode.

Keywords: Port split, Quantum-3

Detected in version:

40.44.0212

4230775

Description: Due to a known issue, telemetry rate must be set to lower than 3 minutes.

Workaround: N/A

Keywords: Telemetry rate

Detected in version:

40.44.0212

4038325 / 
4031430 / 
4038341 / 
4046105

Description: Connecting to systems with NRZ speeds of 1,10, 25, 40, 50, or 100Gb/s is not supported.

Workaround: N/A

Keywords: NRZ, Connectivity

Detected in version:

40.44.0208

4208960

Description: A packet may be parsed incorrectly, if a driver uses the header_length_field_mask when creating a PARSE_GRAPH_NODE object, and the mask value is not composed of continuous bits or does not commence at the least significant bit.

Workaround: Insert the header_length_field_mask with continuous bits and commence at the least significant bit.

Keywords: PARSE GRAPH NODE, Flex Parser

Detected in version:

40.44.0208

4201405

Description: Upgrading to firmware 40.44.0xxx from any previous Engineering Sample version requires power cycling the driver and not just resetting it (using mlxfwreset).  

Workaround: N/A

Keywords: Upgrade, power cycle, reset

Detected in version:

40.44.0208