NVIDIA ConnectX-7 Adapter Cards Firmware Release Notes v28.46.3048

Bug Fixes History

Note

This section includes history of 3 major releases back. For older releases history, please refer to the relevant firmware versions.

Internal Ref.

Issue

4501157 / 4257750

Description: Fixed a critical issue with a live firmware patch.

Keywords: Live firmware patch

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4516394

Description: Fixed an uncleared state caused performance degradation after migration when there were significant differences in resource allocation by ensuring the state is cleared beforehand.

Keywords: Performance

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4286902

Description: Fixed a race condition in DPA process termination during the exception flow, where a failed process could be missed and not reported to the user.

Keywords: DPA

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4420567

Description: Removed an unnecessary and partially incorrect firmware check that blocked valid action list permutations allowed by the PRM. Validation of these permutations remains the responsibility of the software.

Keywords: Header actions

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4443601

Description: Fixed a firmware issue where PXE failed to boot when both LAG ports were up.

Keywords: PXE, LAG

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4443601

Description: Fixed a firmware issue where PXE failed to boot when both LAG ports were up.

Keywords: PXE, LAG

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4475307

Description: Fixed an issue where PCC DCQCN used incorrect parameter values when link speed was 400Gbps or higher.

Keywords: PCC DCQCN, congestion control

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4480427

Description: Fixed incorrect calculation of start address and mode for the CQE buffer in DPA CQ, which could cause CQEs to be written to the wrong address when the buffer is not 4K-aligned and spans a second page boundary.

Keywords: CQ, CQE Buffer, DPA

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4490103

Description: Fixed the restart timing for the OSFP connector at 400 kHz I2C frequency.

Keywords: Restart timing, OSFP, I2C frequency

Discovered in Version: 28.44.1036

Fixed in Release: 28.46.1006

4416919

Description: Updated Diagnostic Counters interface to prevent the following counters from being cleared after read: pcie_link_latency_total_read_packet and pcie_link_latency_total_read_ns.

Keywords: Diagnostic Counters interface

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4520774

Description: Fixed an issue preventing adp_retx profile in the ROCE_ACCL access register from being set when there are outstanding QPs on the PF or VF.

Keywords: ROCE_ACCL access register, QPs, PF, VF

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4403143

Description: Fixed an issue where CREATE_DPA_PROCESS could fail if a DESTROY_DPA_PROCESS (still running during destroy) was executed on a different VHCA. Also addressed a possible failure of CREATE_DPA_PROCESS after FLR.

Keywords: DPA_PROCESS, FLR

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4388371

Description: Fixed an issue where an uninitialized pport in the SLRG command, when using the SMP interface, caused an assertion failure.

Keywords: SLRG, SMP interface, pport

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4531558

Description: Fixed inconsistent LED behavior where the LED color for max speed was yellow and green otherwise, contrary to specification, due to a swapped GPIO mapping between control and PHY LEDs in the INI file.

Keywords: LED

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4470053

Description: Fixed an issue with vQoS parameter configuration to improve latency handling for large messages.

Keywords: vQoS, latency

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4366117

Description: Configuring a small MTU leads to fragmentation of packets critical for the PXE boot process. As a result, the PXE boot filters mistakenly discard these packets, causing the PXE boot to fail.

Keywords: PXE boot filters

Detected in version: 28.45.1020

Fixed in Release: 28.46.1006

4475307

Description: Fixed an issue where PCC DCQCN used incorrect parameter values when link speed was 400Gbps or higher.

Keywords: PCC DCQCN, congestion control.

Detected in version: 28.45.1020

Fixed in Release: 28.46.1006

4486431

Description: Fixed an issue where issuing multiple parallel queries of DPA_THREAD objects with the same object ID could fail.

Keywords: DPA

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

4497103

Description: Fixed the setting of the adaptive retransmission profile.

Keywords: Adaptive retransmission profile

Discovered in Version: 28.45.1020

Fixed in Release: 28.46.1006

Internal Ref.

Issue

4436922

Description: Fixed DC InfiniBand functionality.

Keywords: DC

Detected in version: 28.45.1020

Fixed in Release: 28.45.1200

Internal Ref.

Issue

4241238

Description: Fixed TX timeout issue related to the esw_scheduling QoS feature.

Keywords: esw_scheduling QoS

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4352025

Description: Fixed an out-of-space issue related to writing Type-Length-Value (TLV) entries to reclaim space by removing outdated or irrelevant configuration entries.

Keywords: TLV

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4318537

Description: Fixed an issue where the AI and HAI parameters of the ZTR_RTTCC algorithm, when configured by users, were automatically overwritten upon link speed changes. With this fix, if AI/HAI values were tuned for link speeds other than 100Gb/s, users should now divide those values by (link_speed / 100) to maintain consistent congestion control algorithm behavior.

Keywords: Congestion control, ZTR_RTTCC

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4368450

Description: Fixed an issue where PCC_CNP_COUNT could not be reset using the pcc_counter.sh script in the DOCA tools.

Keywords: PCC

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4360664

Description: Fixed an issue in DOCA Telemetry recovery following a non-graceful abort or driver restart. The problem was related to configuring certain counters during the post-abort flow.

Keywords: DOCA Telemetry recovery

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4311009

Description: Fixed an issue with doorbell recovery that occurred when DCS was active in the system that resulted in the FLR/DESTROY_QP command getting stuck.

Keywords: DCS, doorbell recovery

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4161925

Description: Fixed an issue where the CDB command timeout needed to be increased due to background traffic.

Keywords: CDB

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4295491

Description: Fixed a rare race condition that could incorrectly detect a lack of communication between the ASICs, resulting in module failure and incorrect reporting as unplugged.

Keywords: Race condition

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4342585

Description: Fixed an issue where TLV with OEM priority was incorrectly processed if blocked by MLNX TLV (over_en=0).

Keywords: TLV

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4209790

Description: Enhanced execution time and reduced memory consumption for ETH transport resources (TIR, Transport Domain).

Keywords: ETH transport resources

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4257863

Description: Fixed an issue that could cause the DESTROY_MKEY command to take an excessively long time to execute, with the host driver displaying a "No done completion" message for this command.

Keywords: MKey

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4208995

Description: Fixed a timing race between Rx tuning and CDR locking following exit from electrical idle.

Keywords: Timing race, Rx, CDR

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4239221

Description: Fixed an issue where performing a software reset could cause the device to become inaccessible, requiring a reboot to restore visibility.

Keywords: sw reset

Discovered in Version: 28.43.1014

Fixed in Release: 28.45.1020

4274327

Description: Fixed an issue in the VQoS algorithm related to learning when an element is active and when it begins sending traffic.

Keywords: VQoS algorithm

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4222773

Description: Reduced the bandwidth fluctuation induced by VQoS rate limiting in systems with bellow 350 QPs.

Note: In this release, the relevant change is enabled by default, while in future versions it will be disabled by default and an additional NV configuration will be required to enable it.

Keywords: VQoS

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4274669

Description: Fixed a race condition that could prevent the application from transmitting when VQoS is enabled.

Keywords: VQoS

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4319008

Description: Fixed an issue that caused bandwidth to drop when unbinding multiple VFs with VQoS enabled.

Keywords: VQoS

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

4199274

Description: Fixed an issue where RTT packets with any destination MAC address were incorrectly treated as having a valid destination MAC. The new firmware now discards RTT packets if their destination MAC does not match the port's MAC.

Keywords: RTT, destination MAC

Discovered in Version: 28.44.1036

Fixed in Release: 28.45.1020

Internal Ref.

Issue

4088444 / 4208995

Description: Fixed a timing race between Rx tuning and CDR locking following exit from electrical idle.

Keywords: Timing race, Rx, CDR

Discovered in Version: 28.44.1036

Fixed in Release: 28.44.1206

Internal Ref.

Issue

4319359

Description: Resolved an issue that caused the SLRG register to be unreadable when using the SMP AccessRegister MAD.

Keywords: MADs

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4087432

Description: Increased the RX lossless buffer size to delay the transmission of Pause/PFC frames during NIC congestion.

Keywords: RX lossless buffer size

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4176327

Description: Fixed cable info semaphore deadlock.

Keywords: Cables

Discovered in Version: 28.43.1014

Fixed in Release: 28.44.1036

4179944

Description: Fixed the error handling for the TLV full list, which caused the TLV mechanism to hang.

Keywords: TLV

Discovered in Version: 28.43.1014

Fixed in Release: 28.44.1036

4199196

Description: Fixed the SPDM GET_CERTIFICATE operation to support all certificate chain offsets and chunk sizes.

Keywords: SPDM

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4183928

Description: Fixed an issue in VDPA where destroying a virtq would cause a health buffer syndrome with ext_synd=0x8f33 if the virtq was created without an mkey or with unmanned and mapped mkeys during live migration.

Keywords: VDPA, live migration

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4184904 / 4183908

Description: Fixed an issue where the VDPA feature bits GUEST_TSO4 and GUEST_TSO6 were unexpectedly set by default, leading to traffic interruptions.

Keywords: VDPA, feature cap, GUEST_TSO4, GUEST_TSO6

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4184904

Description: Fixed an issue where the VDPA feature bits GUEST_TSO4 and GUEST_TSO6 were unexpectedly set by default, leading to traffic interruptions.

Keywords: VDPA, feature cap, GUEST_TSO4, GUEST_TSO6

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4184910

Description: Fixed an issue where enabling PCC NP and setting the link type to one port as IB and the other as Ethernet could cause an assert to appear in dmesg with ext_synd 0x8309.

Keywords: PCC NP, port type

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

4133372

Description: Added support for SyncE at 1G link speed.

Keywords: SyncE

Discovered in Version: 28.43.2026

Fixed in Release: 28.44.1036

3661179

Description: Added a new mechanism for allocations and deallocations flows to enhance parallelism.

Keywords: Allocations, deallocations flows

Discovered in Version: 28.39.1002

Fixed in Release: 28.44.1036

Internal Ref.

Issue

4040226

Description: Added a recovery step in case of CQ doorbell getting lost during VF migration.

Keywords: VF migration

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

3988375

Description: PLDM includes port temperature sensor PDR only if an active cable is connected.

Keywords: Temperature sensor PDR

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

3837332

Description: Changed PCIe Geen4/5 new static configuration for VGA gain and CTLE.

Keywords: PCIe, VGA, CTLE

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4120411

Description: Fixed an issue that occasionally caused PTP accuracy degradation for port speed configured to 1G or 10G.

Keywords: PTP

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4134558

Description: Fixed an issue that resulted in MSIx reduction flow triggered with a wrong limitation (the total number of MSIx is reduced from 8k to 4k by mistake) when the dynamic MSIx feature is enabled and virtio emulation is disabled.

Keywords: Dynamic MSIx

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4007123

Description: Fixed lossless packet drops at 400GB 4 lanes when using an optic fiber cable.

Keywords: 400GB, 4 lanes, optic fiber cable

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4014351

Description: Fixed the query for FACTORY default NV configuration values. The firmware always returned the "next" value to be applied.

Keywords: Access register MNVDA, QUERY / SET configurations

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4048886

Description: Fixed an issue related to override TP4 settings.

Keywords: TP4 settings

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4001690

Description: Changed the CTLE and VGA gain for Gen4/5 starting point.

Keywords: CTLE, VGA gain

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4066248

Description: Increased SPDM's RDT value.

Keywords: SPDM

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4092754

Description: Fixed a rare certificate signature verification error.

Keywords: Certificate signature verification

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4003534

Description: Fixed an issue that caused issues during the cables' linkup process after reinserting a module in multi ASIC platforms.

Keywords: Cables

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4041723

Description: Fixed the user_cc_en default value (mlxreg).

Keywords: mlxreg

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

Internal Ref.

Issue

4040226

Description: Added a recovery step in case of CQ doorbell getting lost during VF migration.

Keywords: VF migration

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

3988375

Description: PLDM includes port temperature sensor PDR only if an active cable is connected.

Keywords: Temperature sensor PDR

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

3837332

Description: Changed PCIe Geen4/5 new static configuration for VGA gain and CTLE.

Keywords: PCIe, VGA, CTLE

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4120411

Description: Fixed an issue that occasionally caused PTP accuracy degradation for port speed configured to 1G or 10G.

Keywords: PTP

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4134558

Description: Fixed an issue that resulted in MSIx reduction flow triggered with a wrong limitation (the total number of MSIx is reduced from 8k to 4k by mistake) when the dynamic MSIx feature is enabled and virtio emulation is disabled.

Keywords: Dynamic MSIx

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4007123

Description: Fixed lossless packet drops at 400GB 4 lanes when using an optic fiber cable.

Keywords: 400GB, 4 lanes, optic fiber cable

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4014351

Description: Fixed the query for FACTORY default NV configuration values. The firmware always returned the "next" value to be applied.

Keywords: Access register MNVDA, QUERY / SET configurations

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4048886

Description: Fixed an issue related to override TP4 settings.

Keywords: TP4 settings

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4001690

Description: Changed the CTLE and VGA gain for Gen4/5 starting point.

Keywords: CTLE, VGA gain

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4066248

Description: Increased SPDM's RDT value.

Keywords: SPDM

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4092754

Description: Fixed a rare certificate signature verification error.

Keywords: Certificate signature verification

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4003534

Description: Fixed an issue that caused issues during the cables' linkup process after reinserting a module in multi ASIC platforms.

Keywords: Cables

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

4041723

Description: Fixed the user_cc_en default value (mlxreg).

Keywords: mlxreg

Discovered in Version: 28.42.1000

Fixed in Release: 28.43.1014

© Copyright 2025, NVIDIA. Last updated on Sep 17, 2025.