Hello, struggling people in this thread. I come with more information that may or may not help with this that got me through some tough times with Thunderbolt.
I had a Thunderbolt P2P network up between my Framework and another computer for some data transfer, and after about half an hour TB3/TB4 died completely:
Jan 30 16:55:25 kronk kernel: ucsi_acpi USBC000:00: UCSI_GET_PDOS failed (-5)
Jan 30 16:55:30 kronk kernel: ucsi_acpi USBC000:00: ucsi_handle_connector_change: ACK failed (-110)
Jan 30 16:55:35 kronk kernel: ucsi_acpi USBC000:00: ucsi_handle_connector_change: ACK failed (-110)
Jan 30 16:55:37 kronk kernel: thunderbolt 1-2: new host found, vendor=0x8086 device=0x1
Jan 30 16:55:37 kronk kernel: thunderbolt 1-2: Intel Corp. skully
Jan 30 16:55:41 kronk kernel: ucsi_acpi USBC000:00: ucsi_handle_connector_change: ACK failed (-110)
Jan 30 16:56:30 kronk kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 30 16:56:42 kronk kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 30 16:57:18 kronk kernel: usb 1-4: reset full-speed USB device number 2 using xhci_hcd
Jan 30 16:57:18 kronk kernel: usb 1-4: reset full-speed USB device number 2 using xhci_hcd
Jan 30 16:57:31 kronk kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 30 16:59:07 kronk kernel: TCP: thunderbolt0: Driver has suspect GRO implementation, TCP performance may be compromised.
--- data transfer is happening during this period ---
Jan 30 17:30:12 kronk kernel: thunderbolt 1-2: host disconnected
--- gone forever ---
Bolt wasn’t even showing any probe events, and when plugging/unplugging any previously working Thunderbolt equipment, udev would only show:
KERNEL[74.301922] change /devices/platform/USBC000:00/typec/port3 (typec)
KERNEL[74.301973] remove /devices/platform/USBC000:00/typec/port3/port3-partner (typec)
KERNEL[74.306847] change /devices/platform/USBC000:00/power_supply/ucsi-source-psy-USBC000:004 (power_supply)
UDEV [74.308617] change /devices/platform/USBC000:00/typec/port3 (typec)
UDEV [74.309721] remove /devices/platform/USBC000:00/typec/port3/port3-partner (typec)
UDEV [74.380470] change /devices/platform/USBC000:00/power_supply/ucsi-source-psy-USBC000:004 (power_supply)
KERNEL[81.155683] change /devices/platform/USBC000:00/typec/port3 (typec)
KERNEL[81.155744] add /devices/platform/USBC000:00/typec/port3/port3-partner (typec)
KERNEL[81.160940] change /devices/platform/USBC000:00/power_supply/ucsi-source-psy-USBC000:004 (power_supply)
UDEV [81.162294] change /devices/platform/USBC000:00/typec/port3 (typec)
UDEV [81.163764] add /devices/platform/USBC000:00/typec/port3/port3-partner (typec)
UDEV [81.233849] change /devices/platform/USBC000:00/power_supply/ucsi-source-psy-USBC000:004 (power_supply)
… and the controller didn’t even let the TB3/4 docks charge the laptop - totally dead.
I switched to a Windows install and it was still dead there, so it looks like something tripped somewhere.
What I did do after reading Mario’s comments about the PD controller was re-flash the BIOS with the same version it’s already running (03.03 / 3.3 as reported in Linux). As this re-flashes the PD and EC chips… everything came back to life and all the Thunderbolt docks and peripherals again work perfectly in Linux.
So it looks like it’s possible for the Framework’s ports/controllers to get into a state where it refuses to interact with Thunderbolt anything? But re-flashing the firmware seems to fix it.