Comment # 14 on bug 1130448 from Al Cho

(In reply to Michal Suchanek from comment #10)
> Created attachment 801336 [details]
> btmon output while trying to suspend
> 
> While trying more tests I found it is not 100% reproducible. It seems it's
> likely to trigger when some sound was played recently on the audio device.
> It probably  goes into some powersave state when not used and does not
> prevent the suspend then.

There are no wakeup command on btmon log and seems not really any trouble of
bluetooth.
If it's not 100% reproducible, may you try remove the bt module on your machine
for testing S3 
Base on our experience, it may need more details on kernel log ... or we may
not confirm the BT caused S3 failed.