I get an interrupt storm when running an image based on 4.14 on cyan (more specifically it is a prototype of chromeos-4.14, ie 4.14 with Chrome OS patches on top of it).

[   60.089234] ->handle_irq():  ffffffffbe2f803f, 
[   60.090455] 0xffffffffbf2af380
[   60.090510] handle_bad_irq+0x0/0x2e5
[   60.090522] ->irq_data.chip(): ffffffffbf2af380, 
[   60.090553]    IRQ_NOPROBE set
[   60.090584] ->handle_irq():  ffffffffbe2f803f, 
[   60.090590] handle_bad_irq+0x0/0x2e5
[   60.090596] ->irq_data.chip(): ffffffffbf2af380, 
[   60.090602] 0xffffffffbf2af380
[   60.090608] ->action():           (null)
[   60.090779] handle_bad_irq+0x0/0x2e5

and so on. The output is that messy because there are just too many bad interrupts.

/proc/interrupts shows:

115:      78302      77589      77478      76348  chv-gpio    0
 135:          0          1          0          0  chv-gpio   20  TS3A227E
 171:     196025     196802     196579     197818  chv-gpio    0
 180:          0          0          0          0  chv-gpio    9  ACPI:Event
 182:          3          1          1          5  chv-gpio   11  i8042
 183:          0          0          0          0  chv-gpio   12  ELAN0000:00
 184:          0          0          0          0  chv-gpio   13  ELAN0001:00
 300:          0          0          0          0  chv-gpio   46  max98090_interrupt
 304:          0          0          0          0  chv-gpio   50  INT33BB:01 cd

Notice interrupts 115 and 171; those are the "bad" ones.

Reverting commit https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/pinctrl/intel/pinctrl-cherryview.c?id=845e405e5e6c9dc9ed10306a4b5bfeaefebc2e84 
fixes the interrupt storm problem, but then the keyboard and touchpad do not work.

The same problem is seen on Celes.

Also submitted as https://bugzilla.kernel.org/show_bug.cgi?id=197953.