The problem do persists in kernel 4.19 and other versions. It only depens, if a different driver triggers the interrupts. If so, they are counting very high. So it's possible, that you had none driver in 4.19 using those interrupts and as a consequence, the bug did not trigger.
@Jarkko Nikula: Since you are still replying, could you please try again and further to get the needed docs, as requested by Jean Delvare?
The problem do persists in kernel 4.19 and other versions. It only depens, if a different driver triggers the interrupts. If so, they are counting very high. So it's possible, that you had none driver in 4.19 using those interrupts and as a consequence, the bug did not trigger.
@Jarkko Nikula: Since you are still replying, could you please try again and further to get the needed docs, as requested by Jean Delvare?