This article will explain about below PICE error reported on MX with SCBE2.

SGLINK(RD-err): PCIE error [0x23] at device 

In an MX platform below error msg can be seen:

Oct 15 00:44:06 MX_960 chassisd[68031]: %DAEMON-6: XFCHIP[1]: post init successful xfchip_init_hw

Oct 15 00:44:06 MX_960 chassisd[68031]: %DAEMON-3: XFCHIP_IOERR 1: read 0x484bfc -> 0 failed

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE error [0x23] at device 0, address 0xc84bfc, unexpected TLP, mal-formed TLP, bad RX completion TLP

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE RX TLP[0] data 0x001e001e

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE RX TLP[1] data 0x001e001e

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE RX TLP[2] data 0x001e001e

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE RX TLP[3] data 0x001e001e

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE TX TLP[0] data 0x00000001

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE TX TLP[1] data 0x00000d0f

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE TX TLP[2] data 0x01c84bfc

Oct 15 00:44:06 MX_960 kernel: %KERN-3: SGLINK(RD-err): PCIE TX TLP[3] data 0x00000000

SGBus are low-speed, parallel, and bidirectional buses that are used to carry control information from the Control Board (CB) to the other Field Replaceable Units (FRUs), including the backup CB.

SGLINK is a serialized version of SGBus

In log messages we will see PCIE errors also we will see the CB alarm as below

2 alarms currently active

Alarm time Class Description

2024-10-15 01:34:27 UTC Minor CB 0 Fabric Chip 1 Not Online

2024-10-15 01:34:26 UTC Minor CB 0 Fabric Chip 0 Not Online

• CB reported PCIE Error, PCIE errors are logged when the error check is happening after the IO transaction done between RE and SCB and have received/read corrupted message from the pci error status register.
• RE failed to read CB's XFchip link-type and clocking registers followed by CB HSL channel links read errors etc

This could be a transient hardware issue, need to restart the CB which reporting this error.

If the issue does not resolve with a CB restart, please contact JTAC support.