Comment 2 for bug 1937897

EDIT: To clarify both times these lines come they refer to "gpiochip1" (so not 2 different ones).

I started getting the same 3 lines of errors TWICE at boot, with the now latest official 5.8 HWE kernel. Without "splash", just "quiet" set for GRUB it also shows on my monitors when booting. Honestly seems to slow down boot too (I mean I have 5.5 GB/s R/W NVMes and it can get to GDM in maybe less than 2 seconds).

WHAT HAPPENED:

After my 20.04.2 (HWE) kernel was updated From:

linux-image-5.8.0-59-generic

To:

linux-image-5.8.0-63-generic

If I boot the former (*-59) or any older, I don't see the below errors TWICE. With the latter (*-63) I do. Seems to be little to none information about this. I only find it in the Linux source code. Is it something new in Linux?

Running journalctl -b (now) first 3 lines of errors come after early kernel NVME(s) preparation, but have just checked that once (now). Seems random as the next exact same 3 lines come below something completely different in the log. But again system can boot in 2 secs, however seems slower with these in fact.

$ journalctl -b

RED LINES:
kernel: gpio gpiochip1: (gpio_aaeon): tried to insert a GPIO chip with zero lines
gpiochip_add_data_with_key: GPIOs 0..-1 (gpio_aaeon) failed to register, -22

YELLOW LINE:
gpio-aaeon: probe of gpio-aaeon.0 failed with error -22

... and again later in the log:

RED LINES:
kernel: gpio gpiochip1: (gpio_aaeon): tried to insert a GPIO chip with zero lines
gpiochip_add_data_with_key: GPIOs 0..-1 (gpio_aaeon) failed to register, -22

YELLOW LINE:
gpio-aaeon: probe of gpio-aaeon.0 failed with error -22

---

Apart from that everything works. No new hardware has been added.

Pretty generic system (hefty CPU but still X570 + Zen2):

* ASUS X570 AM4 Motherboard ("Prime Pro")
* Ryzen Zen2 3950X 16C/32T CPU
* ASUS Radeon RX 5500 XT 8 GB OC GPU
* 2 NVMes and 3 HDDs (total 12 TB)
* No special peripherals

Happens really early at boot, so seems low-level.

Is there at least a way to hide these or disable the kernel from trying and failing? Kernel/GRUB parameter? Module parameter? Blacklist a new module? Just running "locate -i gpio", well it seems that both kernels have the same files containing the name "gpio", but haven't compared 100%. No gpio module, service or anything that wasn't there before.