Comment 100 for bug 1752772

Peter Smith - I had the same experience of trying the 4.15.0-33 and it not working for me.
** Try 4.15.0.23 if you would like to be on the 4.15 kernel **

Background:
In my case I am running Mint, but same Ubuntu kernels. It is a desktop PC which is not often suspended so I noticed the problem only a couple of days ago, when already on 4.15.0-30, so tried the 4.15.0-33 which Mint's update manager was offering.. no improvement. The modprobe trick worked and I have a Realtec NIC.

I had seen references to 4.13.0-31 working and also 4.15.0-20, both of which worked for me.

I worked my way up through the following:-
4.15.0-20 resumes
4.15.0-22 resumes
4.15.0-23 resumes
4.15.0-24 does not resume
4.15.0-30 does not resume
4.15.0-33 does not resume
The modprobe treatment worked for the recent -30 and -33, but curiously not on -24. That's probably a symptom of progress.

I thought I experienced a surprising result of going to -24 then back to -23 and the problem reappearing in -23. After a power-off restart all was well again. I had a theory that maybe since all I was doing were OS "restarts" without the PC shutting down the hardware might be holding some state.
I've tried recreating that problem (because it seemed such a dodgy assertion) - but without luck. On the one hand it might be a valid thought that it is very easy to just do restarts between kernels and actually still be holding some unfortunate state in the hardware - OR my bad in accidentally booting the wrong kernel (I've mostly done a uname -a check straight after the boot)

There do seem to have been some inconsistencies in people's experiences, so I wonder if hardware state can be a factor, after all, we are in the area of suspended motherboard h/w
People who know more about Linux initial initialisation of hardware may say I'm talking bo***cks!

Anyway, I'm happy with suspend/resume with 4.15.0-23-generic