Public bug reported:

On the Gigabyte X870 AORUS ELITE WIFI7 motherboard equipped with an AMD
Ryzen 7 9700X processor, native Linux kernels (specifically tested on
7.0.0-38-generic and later 7.0.0-42-generic) completely lack driver
support for the onboard ITE Super I/O hardware monitoring chips. During
hardware probing (sensors-detect), the system identifies unknown chips
with IDs 0x8696 and 0x8883, but sensors cannot read subfeatures (such as
fan1_min, fan1_max, and power caps), returning persistent N/A readouts
and execution errors.

Because the kernel cannot interface with these newer ITE chip IDs, the
motherboard's Embedded Controller (EC) loses continuous communication
with userspace sensors and repeatedly triggers a failsafe routine,
forcing all system and CPU fans to run continuously at maximum speed
(100% duty cycle) as a thermal protective measure.

Full Incident History & Current Status:

1. Kernel Regressions: Kernel 7.0.0-42-generic was identified as
problematic, placed on hold, and purged. The system was reverted to
running on the older 7.0.0-38-generic kernel, but the underlying sensor
communication failure persists on both kernel versions.

2. Failed Driver Compilation Attempts: Multiple attempts to compile and
load community/third-party it87 drivers and DKMS modules via various
GitHub repositories (ucw/it87, a1winston/it87-dkms, and hjlcmb/it87)
continuously failed due to repository redirection/HTTP errors, missing
dependencies/packages in Mint repositories, and authentication prompts.

3. Hardware Mitigation & Ongoing Instability Monitoring: To immediately
suppress the extreme fan noise, a manual "Normal" fan profile was
applied directly in the BIOS (Gigabyte Smart Fan 6). Following a
complete cold shutdown and power cycle (which temporarily cleared the
EC's locked failsafe state), the fan speeds currently appear to adhere
to the hardware BIOS profile. However, the root cause at the kernel
driver level remains entirely unresolved. Because userspace sensor
readouts still output N/A, it is currently unknown whether the hardware-
level BIOS override will hold over time or if the EC will trip back into
maximum failsafe fan speeds during extended runtime or load.

Environment & System Information:

- OS / Release (lsb_release -rd): Linux Mint (Ubuntu-based distribution)

- Kernel Signature (cat /proc/version_signature): Stored in version.log

- Motherboard: Gigabyte X870 AORUS ELITE WIFI7

- Super I/O Chips: ITE IDs 0x8696 and 0x8883

- Processor: AMD Ryzen 7 9700X 8-Core

Attached Logs:

1. version.log (generated via cat /proc/version_signature > version.log)

2. lspci-vnvn.log (generated via sudo lspci -vnvn > lspci-vnvn.log)

Steps to Reproduce:

1. Boot into Linux Mint on a Gigabyte X870 AORUS ELITE WIFI7
motherboard.

2. Run sensors-detect --auto.

3. Observe Super I/O probe results identifying unsupported chip IDs
0x8696 and 0x8883.

4. Run sensors and observe N/A readouts and subfeature read errors
(fan1_min, fan1_max, etc.).

5. Observe the Embedded Controller failing over to 100% max fan speeds
(which currently requires a cold shutdown and manual BIOS curve forcing
to temporarily suppress).

Expected Behavior:
The Linux kernel (or standard hwmon driver stack) should natively support ITE 
0x8696 and 0x8883 chips to provide proper sensor telemetry, allowing controlled 
fan operation without relying on manual BIOS workarounds or risking EC failsafe 
lock-ins

** Affects: linux (Ubuntu)
     Importance: Undecided
         Status: New

** Attachment added: "it87-bug.zip"
   
https://bugs.launchpad.net/bugs/2173636/+attachment/6015821/+files/it87-bug.zip

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2173636

Title:
  ITE Super I/O chip (IDs 0x8696 / 0x8883) unsupported on Gigabyte X870
  AORUS ELITE WIFI7, causing sensor read-out failures and erratic
  maximum fan speed behavior

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2173636/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to