Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

Ver. 1.60(B) contains updates to the component packaging and is functionally equivalent to ver. 1.60. It is not necessary to upgrade with Revision B if a previous component Revision was used to upgrade the firmware to version 1.60.

This revision of the System ROM includes the latest revision of the Intel microcode which, in combination with operating system updates, provides mitigation for Variant 2 of the Side Channel Analysis vulnerability, also known as Spectre. The revision of the microcode included in this System ROM does NOT have issues with more frequent reboots and unpredictable system behavior which impacted the previous Intel microcode which was part of the Spectre Variant 2 mitigation. Additional information is available from Intel’s Security Exploit Newsroom, https://newsroom.intel.com/press-kits/security-exploits-intel-products/.

Firmware Dependencies:

None

Problems Fixed:

Updated the Intel processor microcode to the latest version.

Known Issues:

None

This revision of the System ROM includes the latest revision of the Intel microcode which, in combination with operating system updates, provides mitigation for Variant 2 of the Side Channel Analysis vulnerability, also known as Spectre. The revision of the microcode included in this System ROM does NOT have issues with more frequent reboots and unpredictable system behavior which impacted the previous Intel microcode which was part of the Spectre Variant 2 mitigation. Additional information is available from Intel’s Security Exploit Newsroom, https://newsroom.intel.com/press-kits/security-exploits-intel-products/.

Firmware Dependencies:

None

Problems Fixed:

Updated the Intel processor microcode to the latest version.

Known Issues:

None

• An issue where injecting EEH (Extended Error Handlng) may cause extra Kernel prints, such as: “EEH: Might be infinite loop in mlx5_core driver” was fixed.
• RPM-based OSs with non-default kernels, using repositories after re-creating the installer (using --add-kernel-support) were resulting in improper installation of the drivers.
• A typo in perftest package where help messages wrongly displayed the conversion result between Gb/s and MB/s (20^2 instead of 2^20).
• When one of the LAG slaves went down, LAG deactivation failed, ultimately causing bandwidth degradation.
• The mlx5 driver maintains a subdirectory for every open eth port in /sys/kernel/debug/. For the default network namespace, the sub-directory name is the name of the interface, like "eth8". The new convention for the network interfaces moved to the non-default network namespaces is the interfaces name followed by "@" and the port's PCI ID. For example: "eth8@0000:af:00.3".
• "openibd" script load used to fail when esp6_offload module did not load successfully.
• Added dependency of package mpi-selectors on perl-Getopt-Long system package. On minimal installs of RPM-based OSs, installing mpi-selectros will also install the required system package perl-Getopt-Long.
• During certain rare scenarios, due to Rx page not being replenished, the same page fragment mistakenly became assigned to two different Rx descriptors.
• Driver load issue existed with Errata-kernel on SLES15 SP1.
• An issue in the Hairpin feature which prevented adding hairpin flows using TC tool was fixed.
• WQ (Work Queue) queue flushing was not handled properly in the event of EEH.
• A rare kernel crash scenario when exiting an application that uses RMPP mads intensively.
• A possible kernel crash scenario when AER/slot reset in done in parallel to user space commands execution.
• Added missing ECN (Explicit Congestion Notification) configuration under sysfs for PFs in SwitchDev mode.
• When firmware response time to commands became very long, some commands failed upon timeout. The driver may have then triggered a timeout completion on the wrong entry, leading to a NULL pointer call trace.
• Added driver support for kernels with the old XDP_REDIRECT infrastructure that uses the following NetDev operations: .ndo_xdp_flush and .ndo_xdp_xmit.
• Usage of --excludedocs Open MPI RPM option resulted in the removal of non-documentation related files.
• Disabled automated loading of some modules through udev triggers to preserve the startup process of previous MLNX_OFED versions.
• Reference count (refcount) for RDMA connection ID (cm_id) was not incremented in rdma_resolve_addr() function, resulting in a cm_id use-after-free access. A fix was applied to increment the cm_id refcount
• A race condition which caused kernel panic when moving two ports to SwitchDev mode at the same time.
• Allowed accessing sysfs hardware counters in SwitchDev mode
• Function smp_processor_id() is called in the RX page recycle flow to determine the core to run on. This is intended to run in NAPI context. However, due to a bug in backporting, the RX page recycle was mistakenly called also in the RQ close flow when not needed.
• Port link state was automatically changed (without admin state involvement) to "UP" after reboot.
• ConnectX-3 adapter cards froze when running over SLES 11 OS.
• RDMA CM (Remote Direct Memory Access Communation Manager) connection failed when port space was small.
• Traffic mirroring with OVS (Open VSwitch)offload and non-offload over VxLAN interface is now supported. Note: For kernel 4.9, make sure to use a dedicated OVS version.
• When working with VF LAG while the bond device is in active-active mode, running fwreset would result in unequal traffic on both PFs, and PFs would not reach line rate.
• When bond was created over VF (Virtuaal Function) netdevices in SwitchDev mode, the VF netdevice would be treated as representor netdevice. This caused the mlx5_core driver to crash in case it received netdevice events related to bond device.
• An issue where following a bad affinity occurrence in VF LAG mode, traffic was sent after the port went up/down in the switch.
• Added support for VLAN (Virtual Local Area Network) header rewrite on CentOS 7.2 OS.

• An issue where injecting EEH (Extended Error Handlng) may cause extra Kernel prints, such as: “EEH: Might be infinite loop in mlx5_core driver” was fixed.
• RPM-based OSs with non-default kernels, using repositories after re-creating the installer (using --add-kernel-support) were resulting in improper installation of the drivers.
• A typo in perftest package where help messages wrongly displayed the conversion result between Gb/s and MB/s (20^2 instead of 2^20).
• When one of the LAG slaves went down, LAG deactivation failed, ultimately causing bandwidth degradation.
• The mlx5 driver maintains a subdirectory for every open eth port in /sys/kernel/debug/. For the default network namespace, the sub-directory name is the name of the interface, like "eth8". The new convention for the network interfaces moved to the non-default network namespaces is the interfaces name followed by "@" and the port's PCI ID. For example: "eth8@0000:af:00.3".
• "openibd" script load used to fail when esp6_offload module did not load successfully.
• Added dependency of package mpi-selectors on perl-Getopt-Long system package. On minimal installs of RPM-based OSs, installing mpi-selectros will also install the required system package perl-Getopt-Long.
• During certain rare scenarios, due to Rx page not being replenished, the same page fragment mistakenly became assigned to two different Rx descriptors.
• Driver load issue existed with Errata-kernel on SLES15 SP1.
• An issue in the Hairpin feature which prevented adding hairpin flows using TC tool was fixed.
• WQ (Work Queue) queue flushing was not handled properly in the event of EEH.
• A rare kernel crash scenario when exiting an application that uses RMPP mads intensively.
• A possible kernel crash scenario when AER/slot reset in done in parallel to user space commands execution.
• Added missing ECN (Explicit Congestion Notification) configuration under sysfs for PFs in SwitchDev mode.
• When firmware response time to commands became very long, some commands failed upon timeout. The driver may have then triggered a timeout completion on the wrong entry, leading to a NULL pointer call trace.
• Added driver support for kernels with the old XDP_REDIRECT infrastructure that uses the following NetDev operations: .ndo_xdp_flush and .ndo_xdp_xmit.
• Usage of --excludedocs Open MPI RPM option resulted in the removal of non-documentation related files.
• Disabled automated loading of some modules through udev triggers to preserve the startup process of previous MLNX_OFED versions.
• Reference count (refcount) for RDMA connection ID (cm_id) was not incremented in rdma_resolve_addr() function, resulting in a cm_id use-after-free access. A fix was applied to increment the cm_id refcount
• A race condition which caused kernel panic when moving two ports to SwitchDev mode at the same time.
• Allowed accessing sysfs hardware counters in SwitchDev mode
• Function smp_processor_id() is called in the RX page recycle flow to determine the core to run on. This is intended to run in NAPI context. However, due to a bug in backporting, the RX page recycle was mistakenly called also in the RQ close flow when not needed.
• Port link state was automatically changed (without admin state involvement) to "UP" after reboot.
• ConnectX-3 adapter cards froze when running over SLES 11 OS.
• RDMA CM (Remote Direct Memory Access Communation Manager) connection failed when port space was small.
• Traffic mirroring with OVS (Open VSwitch)offload and non-offload over VxLAN interface is now supported. Note: For kernel 4.9, make sure to use a dedicated OVS version.
• When working with VF LAG while the bond device is in active-active mode, running fwreset would result in unequal traffic on both PFs, and PFs would not reach line rate.
• When bond was created over VF (Virtuaal Function) netdevices in SwitchDev mode, the VF netdevice would be treated as representor netdevice. This caused the mlx5_core driver to crash in case it received netdevice events related to bond device.
• An issue where following a bad affinity occurrence in VF LAG mode, traffic was sent after the port went up/down in the switch.
• Added support for VLAN (Virtual Local Area Network) header rewrite on CentOS 7.2 OS.

This product addresses an issue where an iSCSI portal is discovered with invalid an VLAN ID.

• Temperature sensors logic inside gSEP model and SES database
• When an IOM is pulled the surviving IOM reports false critical temperatures

Please refer to the Release Notes for the complete listing of fixes, enhancements, known issues and work-arounds corresponding to this firmware.

• Fixed an issue where the "Turn off" LED operation momentarily turns on all Identify LED.
• Fixed an issue where an NVME drive was failed after flashing drive firmware.
• Fixed an issue where an operation failure message was observed while deleting a volume with status “Not Available”.
• Fixed an issue that SSAScripting reported incorrect size for some drives greater than 1TB. This did not affect SSACLI.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

• Drive status all shows failed on SL4540 Gen8 H220 [LSI] ILO4 due to  incorrect SCSI command tags.
• Addressed issue where hp-ams may fail to start when U.3 NVMe devices are installed in the system
• The HPE Synergy 6810C 25/50Gb Ethernet Adapter not displaying information in hp-ams is now resolved

• Addressed a small amount of memory leak as documented in the Customer Advisory, a00053824en_us
•  NVMe drive information is now available in the CPQSE.

• PortRcvPkts counter was prevented from being cleared after resetting it.
• The system Timed Out on the configuration cycle of the Virtual Functions (VFs) when more than 10 Virtual Functions performed FLR and the completion Time Out value was configured to a range of less than 16 msec.
• The server hangs and results in NMI when running “mlxfwtop –d mt4103_pci_cr0” while restarting the driver in parallel (from a different thread). In this case, the downstream bridge over the device reported completion timeout error.
• In flow_steering, BMC could not receive a ping over IPV6 after running bmc_reboot.
• While closing the HCA, the RX packet caused bad access to resources that did not exist, and consequently caused the QPCGW or the irisc to get stuck.
• The master SMLID and the LID was either 0 or 0xFFFF when the port was neither active nor armed.
• ibdump could not capture all MADs packets.
• link did not go up after reboot.
• Fixed a rare issue that cause the PCIe configuration cycle that arrived during the time of sw_reset to generate 2 completions.
• Network Controller Sideband Interface (NC-SI) did not work when adding the disable_static_steering_ini field in the ini file, due to memory allocation issue for this field in the scratchpad.

Fixes in version 2.42.5056:

• Fixed an issue that resulted in reading from invalid I/O address on handover from UEFI boot to OS boot, when a port was configured as InfiniBand on a VPI adapter device.

• PortRcvPkts counter was prevented from being cleared after resetting it.
• The system Timed Out on the configuration cycle of the Virtual Functions (VFs) when more than 10 Virtual Functions performed FLR and the completion Time Out value was configured to a range of less than 16 msec.
• The server hangs and results in NMI when running “mlxfwtop –d mt4103_pci_cr0” while restarting the driver in parallel (from a different thread). In this case, the downstream bridge over the device reported completion timeout error.
• In flow_steering, BMC could not receive a ping over IPV6 after running bmc_reboot.
• While closing the HCA, the RX packet caused bad access to resources that did not exist, and consequently caused the QPCGW or the irisc to get stuck.
• The master SMLID and the LID was either 0 or 0xFFFF when the port was neither active nor armed.
• ibdump could not capture all MADs packets.
• link did not go up after reboot.
• Fixed a rare issue that cause the PCIe configuration cycle that arrived during the time of sw_reset to generate 2 completions.
• Network Controller Sideband Interface (NC-SI) did not work when adding the disable_static_steering_ini field in the ini file, due to memory allocation issue for this field in the scratchpad.

Fixes in version 2.42.5056:

• Fixed an issue that resulted in reading from invalid I/O address on handover from UEFI boot to OS boot, when a port was configured as InfiniBand on a VPI adapter device.

• Clearing min_rate on all SR-IOV legacy VFs after setting min_rate to at least one of the VFs did not disable QoS min_rate.
• An empty /usr/src/mlnx-ofa_kernel/default is now no longer created.
• An issue where installing packages through a repository failed after generation of metapackages using --add-kernel-support. The failure occurred due to excessive and incorrect Obsolete headers in those metapackages.Enabled installation of mlnx-ofa_kernel package to /bin/python3 instead of / usr/bin/python3 on RHEL 8.x systems.
• All IPoIB offload packets were wrongly counted as rx_csum_complete. These packets are now identified as rx_csum_unnecessary packets.
• mlx5dv_dr_rule_destroy segmentation fault could happen in rare cases with multiple rules on the same matcher with a different number of actions. This could happen after reusing an already deleted rule memory with less actions.
• An issue of moving to SwitchDev mode after configuring DSCP (Differentiated Service Code Point).
• Global traffic class configuration did not take effect on DC QPs (Dynamically Connected Queue Pairs).
• ib_send_bw traffic frequently dropped to zero when RDMA CM was used, because of incorrect min_rnr_timer setting on the responder side. The min_rnr_timer setting is now aligned with the setting in non-RDMA CM cases.
• Traffic did not pass over VFs with VST QinQ feature is enabled.
• When traffic was sent over Geneve VLAN with Tx VLAN offload enabled and TSO (TCP Segmentation Offload) or Tx csum enabled, traffic could be dropped and not sent to the wire.
• An issue that caused the firmware to restart upon installing mlnxofed-dpdk-upstream-libs package manually.
• Bad utility paths in rdma-core "dracut" hooks on SLES systems, which used to result in the following errors when running "dracut" with the "--add rdma" option. dracut-install:
  • ERROR: installing ‘/usr/libexec/mlx4-setup.sh’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/mlx4-setup.sh dracut-install: ERROR: installing ‘/usr/libexec/rdma-set-sriov-vf’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/rdma-set-sriov-vf
• Traffic class value was not updated in DCT when set via sysfs.
• A doorbell loss issue on AMD platforms with Secure Memory Encryption (SME).
• Registering memory using mlx5dv_devx_umem_reg while forking. Without this fix, applications which use fork() or similar syscalls while using a memory registered with umem_reg could hang due to incorrect physical page mapping. This fix requires setting the IBV_FORK_SAFE environment variable.
• When unbinding the device resulted in the following message being printed to the dmesg: "failed to disable DC tracer"
• Content of file /sys/class/net//statistics/ multicast might have been out of date and might have displayed values lower than the real values.
• ERSPAN (Encapsulated Remote SPAN) protocol was available only when turning off Tx checksum offload.
• A udev script issue which caused non-mellanox devices to be renamed.
• Missing representor statistics when using ifconfig.
• Fixed wrong value of skb mark of received packets on representors.
• An issue which caused second port representors to be named as first port representors.

• Clearing min_rate on all SR-IOV legacy VFs after setting min_rate to at least one of the VFs did not disable QoS min_rate.
• An empty /usr/src/mlnx-ofa_kernel/default is now no longer created.
• An issue where installing packages through a repository failed after generation of metapackages using --add-kernel-support. The failure occurred due to excessive and incorrect Obsolete headers in those metapackages.Enabled installation of mlnx-ofa_kernel package to /bin/python3 instead of / usr/bin/python3 on RHEL 8.x systems.
• All IPoIB offload packets were wrongly counted as rx_csum_complete. These packets are now identified as rx_csum_unnecessary packets.
• mlx5dv_dr_rule_destroy segmentation fault could happen in rare cases with multiple rules on the same matcher with a different number of actions. This could happen after reusing an already deleted rule memory with less actions.
• An issue of moving to SwitchDev mode after configuring DSCP (Differentiated Service Code Point).
• Global traffic class configuration did not take effect on DC QPs (Dynamically Connected Queue Pairs).
• ib_send_bw traffic frequently dropped to zero when RDMA CM was used, because of incorrect min_rnr_timer setting on the responder side. The min_rnr_timer setting is now aligned with the setting in non-RDMA CM cases.
• Traffic did not pass over VFs with VST QinQ feature is enabled.
• When traffic was sent over Geneve VLAN with Tx VLAN offload enabled and TSO (TCP Segmentation Offload) or Tx csum enabled, traffic could be dropped and not sent to the wire.
• An issue that caused the firmware to restart upon installing mlnxofed-dpdk-upstream-libs package manually.
• Bad utility paths in rdma-core "dracut" hooks on SLES systems, which used to result in the following errors when running "dracut" with the "--add rdma" option. dracut-install:
  • ERROR: installing ‘/usr/libexec/mlx4-setup.sh’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/mlx4-setup.sh dracut-install: ERROR: installing ‘/usr/libexec/rdma-set-sriov-vf’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/rdma-set-sriov-vf
• Traffic class value was not updated in DCT when set via sysfs.
• A doorbell loss issue on AMD platforms with Secure Memory Encryption (SME).
• Registering memory using mlx5dv_devx_umem_reg while forking. Without this fix, applications which use fork() or similar syscalls while using a memory registered with umem_reg could hang due to incorrect physical page mapping. This fix requires setting the IBV_FORK_SAFE environment variable.
• When unbinding the device resulted in the following message being printed to the dmesg: "failed to disable DC tracer"
• Content of file /sys/class/net//statistics/ multicast might have been out of date and might have displayed values lower than the real values.
• ERSPAN (Encapsulated Remote SPAN) protocol was available only when turning off Tx checksum offload.
• A udev script issue which caused non-mellanox devices to be renamed.
• Missing representor statistics when using ifconfig.
• Fixed wrong value of skb mark of received packets on representors.
• An issue which caused second port representors to be named as first port representors.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

• An issue that caused the DCR (DC Resources) to be destroyed before the retry option managed to work when the retry timeout is too big. in this case, the DCR' time-to-live was increased, and the maximum retry timeout was decreased.
• Increased PHY power consumption limit to 1.5w.
• An issue that caused PortCounters.PortRcvErr / PPCNT.infiniband_counters.PortRcvErr not to report port icrc errors.

The following issues have been fixed in version 16.31.1014:

• TX PRBS was not changed even after reconfiguring it, all PRBS mode were enabled in test mode.
• An assert that was caused when trying to open 1024 functions on the device and the maximum number of functions was 1023.
• Occasional performance issues related to RC QPs using E2E-credits (not connected to SRQ and doing send/receive traffic) when the ROCE_ACCL tx_window was enabled.
• A fatal error occurred and eventually resulted in the HCA to get into an unresponsive state when a packet was larger than a strided receive WQE that was being scattered.
• A rare issue that caused RX pipe to hang.
• The resource number size (a 64 bit number) was overwritten with a 32 bit number and erased the high bits when de-allocating the resource number.
• CRT_DCR with index larger than 1 << 21 occasionally collided with the CRT_SW_RESERVED address.
• An issue that caused the TX queue to get into an unresponsive state when the VF rate limiter was set.

Additional Fixes included in version 16.31.1014:

• Disabled the CNP counter “rp_cnp_ignored " (triggered by OOS (out-of-sequence)) when all ports are IB.
• Fixed the TMP421 sensor temperature reporting.
• Fixed the rate select mechanism in QSFP modules.
• Fixed classification issues for "Passive" cables to be more robust.
• Initialized the rate table in the static configuration so it will be configured at the link-not-up scenarios.

• An issue that prevented VXLAN packets with svlan/ cvlan tag from being matched.
• An issue that caused the eth_wqe_too_small counter to count ODP (On-Demand Paging) page faults.
• When MKEY_BY_NAME was enabled by NVCONFIG and a large number of VFs are configured, VM restart (VF/PF FLR) will take longer than when MKEY_BY_NAME is disabled.
•  Ani ssue that resulted in low performance after enabling the RoCE Accelerator capability. Note: The fix is available when all ports are set as Ethernet. 
• On rare cases, a fatal error related to errors from the PCI transport layer might be reported during FLR.
• The chassis manager calculation for Multi-Host and Socket-Direct adapter cards to allow running NC-SI commands by the chassis manager BMC. Now the chassis manager is count as BMC with index 0, regardless of how many BMC there are.
• An issue that caused the device to go to down IRISC as one of the firmware semaphores could not be released when a speed change or port state change was triggered. 

Following issues have been fixed in firmware version 12.28.1002:

• An issue that caused the DCR (DC Resources) to be destroyed before the retry option managed to work when the retry timeout is too big. in this case, the DCR' time-to-live was increased, and the maximum retry timeout was decreased.
• Increased PHY power consumption limit to 1.5w.
• An issue that caused PortCounters.PortRcvErr / PPCNT.infiniband_counters.PortRcvErr not to report port icrc errors.

• An issue where injecting EEH (Extended Error Handlng) may cause extra Kernel prints, such as: “EEH: Might be infinite loop in mlx5_core driver” was fixed.
• RPM-based OSs with non-default kernels, using repositories after re-creating the installer (using --add-kernel-support) were resulting in improper installation of the drivers.
• A typo in perftest package where help messages wrongly displayed the conversion result between Gb/s and MB/s (20^2 instead of 2^20).
• When one of the LAG slaves went down, LAG deactivation failed, ultimately causing bandwidth degradation.
• The mlx5 driver maintains a subdirectory for every open eth port in /sys/kernel/debug/. For the default network namespace, the sub-directory name is the name of the interface, like "eth8". The new convention for the network interfaces moved to the non-default network namespaces is the interfaces name followed by "@" and the port's PCI ID. For example: "eth8@0000:af:00.3".
• "openibd" script load used to fail when esp6_offload module did not load successfully.
• Added dependency of package mpi-selectors on perl-Getopt-Long system package. On minimal installs of RPM-based OSs, installing mpi-selectros will also install the required system package perl-Getopt-Long.
• During certain rare scenarios, due to Rx page not being replenished, the same page fragment mistakenly became assigned to two different Rx descriptors.
• Driver load issue existed with Errata-kernel on SLES15 SP1.
• An issue in the Hairpin feature which prevented adding hairpin flows using TC tool was fixed.
• WQ (Work Queue) queue flushing was not handled properly in the event of EEH.
• A rare kernel crash scenario when exiting an application that uses RMPP mads intensively.
• A possible kernel crash scenario when AER/slot reset in done in parallel to user space commands execution.
• Added missing ECN (Explicit Congestion Notification) configuration under sysfs for PFs in SwitchDev mode.
• When firmware response time to commands became very long, some commands failed upon timeout. The driver may have then triggered a timeout completion on the wrong entry, leading to a NULL pointer call trace.
• Added driver support for kernels with the old XDP_REDIRECT infrastructure that uses the following NetDev operations: .ndo_xdp_flush and .ndo_xdp_xmit.
• Usage of --excludedocs Open MPI RPM option resulted in the removal of non-documentation related files.
• Disabled automated loading of some modules through udev triggers to preserve the startup process of previous MLNX_OFED versions.
• Reference count (refcount) for RDMA connection ID (cm_id) was not incremented in rdma_resolve_addr() function, resulting in a cm_id use-after-free access. A fix was applied to increment the cm_id refcount
• A race condition which caused kernel panic when moving two ports to SwitchDev mode at the same time.
• Allowed accessing sysfs hardware counters in SwitchDev mode
• Function smp_processor_id() is called in the RX page recycle flow to determine the core to run on. This is intended to run in NAPI context. However, due to a bug in backporting, the RX page recycle was mistakenly called also in the RQ close flow when not needed.
• Port link state was automatically changed (without admin state involvement) to "UP" after reboot.
• ConnectX-3 adapter cards froze when running over SLES 11 OS.
• RDMA CM (Remote Direct Memory Access Communation Manager) connection failed when port space was small.
• Traffic mirroring with OVS (Open VSwitch)offload and non-offload over VxLAN interface is now supported. Note: For kernel 4.9, make sure to use a dedicated OVS version.
• When working with VF LAG while the bond device is in active-active mode, running fwreset would result in unequal traffic on both PFs, and PFs would not reach line rate.
• When bond was created over VF (Virtuaal Function) netdevices in SwitchDev mode, the VF netdevice would be treated as representor netdevice. This caused the mlx5_core driver to crash in case it received netdevice events related to bond device.
• An issue where following a bad affinity occurrence in VF LAG mode, traffic was sent after the port went up/down in the switch.
• Added support for VLAN (Virtual Local Area Network) header rewrite on CentOS 7.2 OS.

• An issue where injecting EEH (Extended Error Handlng) may cause extra Kernel prints, such as: “EEH: Might be infinite loop in mlx5_core driver” was fixed.
• RPM-based OSs with non-default kernels, using repositories after re-creating the installer (using --add-kernel-support) were resulting in improper installation of the drivers.
• A typo in perftest package where help messages wrongly displayed the conversion result between Gb/s and MB/s (20^2 instead of 2^20).
• When one of the LAG slaves went down, LAG deactivation failed, ultimately causing bandwidth degradation.
• The mlx5 driver maintains a subdirectory for every open eth port in /sys/kernel/debug/. For the default network namespace, the sub-directory name is the name of the interface, like "eth8". The new convention for the network interfaces moved to the non-default network namespaces is the interfaces name followed by "@" and the port's PCI ID. For example: "eth8@0000:af:00.3".
• "openibd" script load used to fail when esp6_offload module did not load successfully.
• Added dependency of package mpi-selectors on perl-Getopt-Long system package. On minimal installs of RPM-based OSs, installing mpi-selectros will also install the required system package perl-Getopt-Long.
• During certain rare scenarios, due to Rx page not being replenished, the same page fragment mistakenly became assigned to two different Rx descriptors.
• Driver load issue existed with Errata-kernel on SLES15 SP1.
• An issue in the Hairpin feature which prevented adding hairpin flows using TC tool was fixed.
• WQ (Work Queue) queue flushing was not handled properly in the event of EEH.
• A rare kernel crash scenario when exiting an application that uses RMPP mads intensively.
• A possible kernel crash scenario when AER/slot reset in done in parallel to user space commands execution.
• Added missing ECN (Explicit Congestion Notification) configuration under sysfs for PFs in SwitchDev mode.
• When firmware response time to commands became very long, some commands failed upon timeout. The driver may have then triggered a timeout completion on the wrong entry, leading to a NULL pointer call trace.
• Added driver support for kernels with the old XDP_REDIRECT infrastructure that uses the following NetDev operations: .ndo_xdp_flush and .ndo_xdp_xmit.
• Usage of --excludedocs Open MPI RPM option resulted in the removal of non-documentation related files.
• Disabled automated loading of some modules through udev triggers to preserve the startup process of previous MLNX_OFED versions.
• Reference count (refcount) for RDMA connection ID (cm_id) was not incremented in rdma_resolve_addr() function, resulting in a cm_id use-after-free access. A fix was applied to increment the cm_id refcount
• A race condition which caused kernel panic when moving two ports to SwitchDev mode at the same time.
• Allowed accessing sysfs hardware counters in SwitchDev mode
• Function smp_processor_id() is called in the RX page recycle flow to determine the core to run on. This is intended to run in NAPI context. However, due to a bug in backporting, the RX page recycle was mistakenly called also in the RQ close flow when not needed.
• Port link state was automatically changed (without admin state involvement) to "UP" after reboot.
• ConnectX-3 adapter cards froze when running over SLES 11 OS.
• RDMA CM (Remote Direct Memory Access Communation Manager) connection failed when port space was small.
• Traffic mirroring with OVS (Open VSwitch)offload and non-offload over VxLAN interface is now supported. Note: For kernel 4.9, make sure to use a dedicated OVS version.
• When working with VF LAG while the bond device is in active-active mode, running fwreset would result in unequal traffic on both PFs, and PFs would not reach line rate.
• When bond was created over VF (Virtuaal Function) netdevices in SwitchDev mode, the VF netdevice would be treated as representor netdevice. This caused the mlx5_core driver to crash in case it received netdevice events related to bond device.
• An issue where following a bad affinity occurrence in VF LAG mode, traffic was sent after the port went up/down in the switch.
• Added support for VLAN (Virtual Local Area Network) header rewrite on CentOS 7.2 OS.

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

The following issues have been fixed in version 14.31.1200:

• TX PRBS was not changed even after reconfiguring it, all PRBS mode were enabled in test mode.
• A fatal error occurred and eventually resulted in the HCA to get into an unresponsive state when a packet was larger than a strided receive WQE that was being scattered.
• The resource number size (a 64 bit number) was overwritten with a 32 bit number and erased the high bits when de-allocating the resource number.

Additional Fixes included in version 14.31.1200:

• Fixed the rate select mechanism in QSFP modules.
• Fixed classification issues for "Passive" cables to be more robust.
• Initialized the rate table in the static configuration so it will be configured at the link-not-up scenarios.

The following issues have been fixed in version 16.31.1014:

• TX PRBS was not changed even after reconfiguring it, all PRBS mode were enabled in test mode.
• An assert that was caused when trying to open 1024 functions on the device and the maximum number of functions was 1023.
• Occasional performance issues related to RC QPs using E2E-credits (not connected to SRQ and doing send/receive traffic) when the ROCE_ACCL tx_window was enabled.
• A fatal error occurred and eventually resulted in the HCA to get into an unresponsive state when a packet was larger than a strided receive WQE that was being scattered.
• A rare issue that caused RX pipe to hang.
• The resource number size (a 64 bit number) was overwritten with a 32 bit number and erased the high bits when de-allocating the resource number.
• CRT_DCR with index larger than 1 << 21 occasionally collided with the CRT_SW_RESERVED address.
• An issue that caused the TX queue to get into an unresponsive state when the VF rate limiter was set.

Additional Fixes included in version 16.31.1014:

• Disabled the CNP counter “rp_cnp_ignored " (triggered by OOS (out-of-sequence)) when all ports are IB.
• Fixed the TMP421 sensor temperature reporting.
• Fixed the rate select mechanism in QSFP modules.
• Fixed classification issues for "Passive" cables to be more robust.
• Initialized the rate table in the static configuration so it will be configured at the link-not-up scenarios.

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

Fixes included in firmware version 14.29.1016:

• Changed the default value of DCQCN's (Data center- Quantized Congestion Notification)  NP parameter min_time_between_cnps to 4 on all devices to support larger scalability of cluster.
• An issue that caused packets to drop due to header size issues and/or failing checks. The issue was caused due to a Linux issue that caused VFs to set the wrong header size value in wqe_inline_header_mode input.
• When MKEY_BY_NAME was enabled by NVCONFIG and a large number of VFs were configured, VM restart (VF/PF FLR) will take longer than when MKEY_BY_NAME is disabled.
• On rare cases, a fatal error related to errors from the PCI transport layer might be reported during FLR.
•  An issue that caused the device to go to dead IRISC as one of the firmware semaphores could not be released when a speed change or port state change was triggered. 

No fixes are included in firmware version 14.31.1200.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

Please reference Customer Advisory a00095304en_us.

Please reference Customer Advisory SIK5264.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

This product addresses an issue where the ixgben driver has high CPU overhead when an SFP+ module is absent.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Ver. 2.80_10-16-2020 (B)  contains updates to the component packaging and is functionally equivalent to ver. 2.80_10-16-2020.  It is not necessary to upgrade with Revision B if a previous component Revision was used to upgrade the system ROM to version 2.80_10-16-2020.

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

• FW Transfer Fix
• Drive Failure Fix
• Power Loss Protection
• FW Download Fix
• Reset Fix
• Shutdown fix
• PCIe Improvement
• Random Write Improvement

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

The following issues have been fixed in version 14.31.1200:

• TX PRBS was not changed even after reconfiguring it, all PRBS mode were enabled in test mode.
• A fatal error occurred and eventually resulted in the HCA to get into an unresponsive state when a packet was larger than a strided receive WQE that was being scattered.
• The resource number size (a 64 bit number) was overwritten with a 32 bit number and erased the high bits when de-allocating the resource number.

Additional Fixes included in version 14.31.1200:

• Fixed the rate select mechanism in QSFP modules.
• Fixed classification issues for "Passive" cables to be more robust.
• Initialized the rate table in the static configuration so it will be configured at the link-not-up scenarios.

The following issues have been fixed in version 16.31.1014:

• TX PRBS was not changed even after reconfiguring it, all PRBS mode were enabled in test mode.
• An assert that was caused when trying to open 1024 functions on the device and the maximum number of functions was 1023.
• Occasional performance issues related to RC QPs using E2E-credits (not connected to SRQ and doing send/receive traffic) when the ROCE_ACCL tx_window was enabled.
• A fatal error occurred and eventually resulted in the HCA to get into an unresponsive state when a packet was larger than a strided receive WQE that was being scattered.
• A rare issue that caused RX pipe to hang.
• The resource number size (a 64 bit number) was overwritten with a 32 bit number and erased the high bits when de-allocating the resource number.
• CRT_DCR with index larger than 1 << 21 occasionally collided with the CRT_SW_RESERVED address.
• An issue that caused the TX queue to get into an unresponsive state when the VF rate limiter was set.

Additional Fixes included in version 16.31.1014:

• Disabled the CNP counter “rp_cnp_ignored " (triggered by OOS (out-of-sequence)) when all ports are IB.
• Fixed the TMP421 sensor temperature reporting.
• Fixed the rate select mechanism in QSFP modules.
• Fixed classification issues for "Passive" cables to be more robust.
• Initialized the rate table in the static configuration so it will be configured at the link-not-up scenarios.

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

Fixes included in firmware version 14.29.1016:

• Changed the default value of DCQCN's (Data center- Quantized Congestion Notification)  NP parameter min_time_between_cnps to 4 on all devices to support larger scalability of cluster.
• An issue that caused packets to drop due to header size issues and/or failing checks. The issue was caused due to a Linux issue that caused VFs to set the wrong header size value in wqe_inline_header_mode input.
• When MKEY_BY_NAME was enabled by NVCONFIG and a large number of VFs were configured, VM restart (VF/PF FLR) will take longer than when MKEY_BY_NAME is disabled.
• On rare cases, a fatal error related to errors from the PCI transport layer might be reported during FLR.
•  An issue that caused the device to go to dead IRISC as one of the firmware semaphores could not be released when a speed change or port state change was triggered. 

No fixes are included in firmware version 14.31.1200.

• TBD

• Addressed issue where the HPE Synergy 3820C 10/20Gb Converged Network Adapter was showing not supported
• Corrected invalid utf8mb4 char in database

• Clearing min_rate on all SR-IOV legacy VFs after setting min_rate to at least one of the VFs did not disable QoS min_rate.
• An empty /usr/src/mlnx-ofa_kernel/default is now no longer created.
• An issue where installing packages through a repository failed after generation of metapackages using --add-kernel-support. The failure occurred due to excessive and incorrect Obsolete headers in those metapackages.Enabled installation of mlnx-ofa_kernel package to /bin/python3 instead of / usr/bin/python3 on RHEL 8.x systems.
• All IPoIB offload packets were wrongly counted as rx_csum_complete. These packets are now identified as rx_csum_unnecessary packets.
• mlx5dv_dr_rule_destroy segmentation fault could happen in rare cases with multiple rules on the same matcher with a different number of actions. This could happen after reusing an already deleted rule memory with less actions.
• An issue of moving to SwitchDev mode after configuring DSCP (Differentiated Service Code Point).
• Global traffic class configuration did not take effect on DC QPs (Dynamically Connected Queue Pairs).
• ib_send_bw traffic frequently dropped to zero when RDMA CM was used, because of incorrect min_rnr_timer setting on the responder side. The min_rnr_timer setting is now aligned with the setting in non-RDMA CM cases.
• Traffic did not pass over VFs with VST QinQ feature is enabled.
• When traffic was sent over Geneve VLAN with Tx VLAN offload enabled and TSO (TCP Segmentation Offload) or Tx csum enabled, traffic could be dropped and not sent to the wire.
• An issue that caused the firmware to restart upon installing mlnxofed-dpdk-upstream-libs package manually.
• Bad utility paths in rdma-core "dracut" hooks on SLES systems, which used to result in the following errors when running "dracut" with the "--add rdma" option. dracut-install:
  • ERROR: installing ‘/usr/libexec/mlx4-setup.sh’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/mlx4-setup.sh dracut-install: ERROR: installing ‘/usr/libexec/rdma-set-sriov-vf’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/rdma-set-sriov-vf
• Traffic class value was not updated in DCT when set via sysfs.
• A doorbell loss issue on AMD platforms with Secure Memory Encryption (SME).
• Registering memory using mlx5dv_devx_umem_reg while forking. Without this fix, applications which use fork() or similar syscalls while using a memory registered with umem_reg could hang due to incorrect physical page mapping. This fix requires setting the IBV_FORK_SAFE environment variable.
• When unbinding the device resulted in the following message being printed to the dmesg: "failed to disable DC tracer"
• Content of file /sys/class/net//statistics/ multicast might have been out of date and might have displayed values lower than the real values.
• ERSPAN (Encapsulated Remote SPAN) protocol was available only when turning off Tx checksum offload.
• A udev script issue which caused non-mellanox devices to be renamed.
• Missing representor statistics when using ifconfig.
• Fixed wrong value of skb mark of received packets on representors.
• An issue which caused second port representors to be named as first port representors.

• Clearing min_rate on all SR-IOV legacy VFs after setting min_rate to at least one of the VFs did not disable QoS min_rate.
• An empty /usr/src/mlnx-ofa_kernel/default is now no longer created.
• An issue where installing packages through a repository failed after generation of metapackages using --add-kernel-support. The failure occurred due to excessive and incorrect Obsolete headers in those metapackages.Enabled installation of mlnx-ofa_kernel package to /bin/python3 instead of / usr/bin/python3 on RHEL 8.x systems.
• All IPoIB offload packets were wrongly counted as rx_csum_complete. These packets are now identified as rx_csum_unnecessary packets.
• mlx5dv_dr_rule_destroy segmentation fault could happen in rare cases with multiple rules on the same matcher with a different number of actions. This could happen after reusing an already deleted rule memory with less actions.
• An issue of moving to SwitchDev mode after configuring DSCP (Differentiated Service Code Point).
• Global traffic class configuration did not take effect on DC QPs (Dynamically Connected Queue Pairs).
• ib_send_bw traffic frequently dropped to zero when RDMA CM was used, because of incorrect min_rnr_timer setting on the responder side. The min_rnr_timer setting is now aligned with the setting in non-RDMA CM cases.
• Traffic did not pass over VFs with VST QinQ feature is enabled.
• When traffic was sent over Geneve VLAN with Tx VLAN offload enabled and TSO (TCP Segmentation Offload) or Tx csum enabled, traffic could be dropped and not sent to the wire.
• An issue that caused the firmware to restart upon installing mlnxofed-dpdk-upstream-libs package manually.
• Bad utility paths in rdma-core "dracut" hooks on SLES systems, which used to result in the following errors when running "dracut" with the "--add rdma" option. dracut-install:
  • ERROR: installing ‘/usr/libexec/mlx4-setup.sh’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/mlx4-setup.sh dracut-install: ERROR: installing ‘/usr/libexec/rdma-set-sriov-vf’ dracut: /usr/lib/dracut/dracut-install -D /var/tmp/dracut.UdCOSJ/ initramfs /usr/libexec/rdma-set-sriov-vf
• Traffic class value was not updated in DCT when set via sysfs.
• A doorbell loss issue on AMD platforms with Secure Memory Encryption (SME).
• Registering memory using mlx5dv_devx_umem_reg while forking. Without this fix, applications which use fork() or similar syscalls while using a memory registered with umem_reg could hang due to incorrect physical page mapping. This fix requires setting the IBV_FORK_SAFE environment variable.
• When unbinding the device resulted in the following message being printed to the dmesg: "failed to disable DC tracer"
• Content of file /sys/class/net//statistics/ multicast might have been out of date and might have displayed values lower than the real values.
• ERSPAN (Encapsulated Remote SPAN) protocol was available only when turning off Tx checksum offload.
• A udev script issue which caused non-mellanox devices to be renamed.
• Missing representor statistics when using ifconfig.
• Fixed wrong value of skb mark of received packets on representors.
• An issue which caused second port representors to be named as first port representors.

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

Following issues have been fixed in firmware version 14.29.1016:

• The sent packet hung while the device entered FLR mode.
• Enabled Bar configuration bytewise by applying the write_en bitmask.
• The device hung while running the sw reset flow under heavy stress and with many open resources.
• The completions (CQ)completed due to a race condition in the firmware transport error handlers, and the error stressors, where the error stressors would hang the firmware transport error handler flow. 
• The fragmented IP packets were dropped.

Following issues have been fixed in firmware version 16.29.1016:

• Although the effective BER (after FEC) was expected to meet the design targets (e.g. 10e-14 or lower), occasionally it was higher.
• High BER occured when connecting cables of type 0.5/1m DAC to an HDR speed.
• The PCIe Tx parameters did not load correctly when the speed was changed after the PCIe link was disabled.
• The desched_threshold field did not work properly.
• The "roce_adp_retrans" counter was presenting the values of the "local_ack_timeout_err" counter.
• If Relaxed Ordering was disabled by running the "setpci" command, it would not be functional even after re-enabling it by running the "setpci" command again.

• An issue that prevented the firmware from detecting a link_down event thus preventing the IB bond interface from going to a failover mode.

Following issues have been fixed in firmware version 14.27.4000:

• The sent packet hung while the device entered FLR mode.
• Enabled Bar configuration bytewise by applying the write_en bitmask.
• The device hung while running the sw reset flow under heavy stress and with many open resources.
• The completions (CQ)completed due to a race condition in the firmware transport error handlers, and the error stressors, where the error stressors would hang the firmware transport error handler flow. 
• The fragmented IP packets were dropped.

Following issues have been fixed in firmware version 16.27.2008:

• Although the effective BER (after FEC) was expected to meet the design targets (e.g. 10e-14 or lower), occasionally it was higher.
• High BER occured when connecting cables of type 0.5/1m DAC to an HDR speed.
• The PCIe Tx parameters did not load correctly when the speed was changed after the PCIe link was disabled.
• The desched_threshold field did not work properly.
• The "roce_adp_retrans" counter was presenting the values of the "local_ack_timeout_err" counter.
• If Relaxed Ordering was disabled by running the "setpci" command, it would not be functional even after re-enabling it by running the "setpci" command again.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for security vulnerabilities documented as CVE-2020-24512. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-464. The Intel microcode patches included in this release are version 0x000000EA (CPUID 506E3 and CPUID 906E9). These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel sighting CVE-2020-24512. This security vulnerability is documented in Intel Security Advisory INTEL-SA-00463. The Intel microcode patches included in this release are version 0x000000EA (CPUID 506E3) and 0x000000EA (CPUID 906E9). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-8670 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This revision of the System ROM includes the latest revision of Intel microcode.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode: 0x000000E2 (CPUID 506E3). This Intel microcode patch provide mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-8696 and CVE-2020-8698. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00381. These issues are not unique to HPE servers.

Known Issues:

None

None

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigation for an Intel sighting where the system may experience a machine check after updating to the latest System ROM which contained a fix for an Intel TSX (Transactional Synchronizations Extensions) sightings. The previous microcode was first introduced in the v2.62 System ROM. This issue only impacts systems configured with Intel Xeon® D-1500 processors. This issue is not unique to HPE servers.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

• FW Transfer Fix
• Drive Failure Fix
• Power Loss Protection
• FW Download Fix
• Reset Fix
• Shutdown fix
• PCIe Improvement
• Random Write Improvement

• The "esxcli network sriovnic vf stats" command was not supported. When running this command on a vmknic, a failure message was displayed.

• The "esxcli network sriovnic vf stats" command was not supported. When running this command on a vmknic, a failure message was displayed.

This product addresses an issue where the ixgben driver has high CPU overhead when an SFP+ module is absent.

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2020.2 guidance.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode. The Intel microcode patches included in this release are versions 0x00000044 (CPUID 306F2) and 0x00B00003A (CPUID 406F1).

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for security vulnerabilities documented as CVE-2020-24512. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-464. The Intel microcode patches included in this release are version 0x000000EA (CPUID 506E3 and CPUID 906E9). These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel sighting CVE-2020-24512. This security vulnerability is documented in Intel Security Advisory INTEL-SA-00463. The Intel microcode patches included in this release are version 0x000000EA (CPUID 506E3) and 0x000000EA (CPUID 906E9). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-8670 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

This revision of the System ROM includes the latest revision of Intel microcode.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of Intel microcode: 0x000000E2 (CPUID 506E3). This Intel microcode patch provide mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-8696 and CVE-2020-8698. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00381. These issues are not unique to HPE servers.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-0591 and CVE-2020-0592. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00358. These issues are not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS platform code advisories and security vulnerabilities documented as CVE-2020-8738, CVE-2020-8740, and CVE-2020-8764. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00390. These issues are not unique to HPE servers.

This version of the System ROM contains an updated UEFI Driver for the SmartArray B140i SATA RAID controller. This update addresses an issue where the system may periodically experience an operating system boot issue such as experiencing a Linux kernel panic on boot.

Known Issues:

This version of the System ROM contains updates aligned with the Intel Product Update (IPU) version IPU.2021.1 guidance.

Ver. 2.90_04-29-2021 (B) contains support for Windows Server 2019.  It is functionally equivalent to ver. 2.90_04-29-2021.  It is not necessary to upgrade with version 2.90_04-29-2021 (B) if a previous component revision was used to upgrade the firmware to ver. 2.90_04-29-2021.

Firmware Dependencies:

None

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel microcode which provides mitigations for the Intel technical sighting referred to as MD Clear. The Intel microcode patches included in this release are version 0x00000046 (CPUID 306F2) and 0x0B00003E (CPUID 406F1). This issue is not unique to HPE servers.

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

None

Firmware Dependencies:

Problems Fixed:

This revision of the System ROM includes the latest revision of the Intel Reference Code which provides mitigations for BIOS advisories and security vulnerabilities documented as CVE-2020-12360 and CVE-2020-12357. These security vulnerabilities are documented in Intel Security Advisory INTEL-SA-00463. These issues are not unique to HPE servers.

Known Issues:

-Removed action over FAULT_SENSED bit due to incorrect algorithm.

-Removed action over FAULT_SENSED bit due to incorrect algorithm.

• Due to a SLES 15 kernel setting, hfi1_eprom cannot work while the HFI driver is loaded. The tool and driver are mutually exclusive.