Hmm, in the log you attached, irq 40 is assigned to the HD-audio device, so I guess it's a different situation. Could you try to reproduce the bug and get hwinfo / dmesg outputs from that state?