Holger, On Wed, 7 Nov 2018, Thomas Gleixner wrote: > On Wed, 7 Nov 2018, Holger Schurig wrote: > > > > Though SERIRQ is a really good hint. Let me look into the details of that a > > > bit more. > > > > Maybe I have more hints to you: > > > > - nolapic=1 also helps > > Yeah, that's not surprising. Can you please provide /proc/interrupts for > that with i915 module loaded ? > > > - preventing i915.ko from being loaded also helps > > That's interesting indeed. Btw, the off by one is a red herring, but still > impressive detective work! > > > vector 44 is audio and vector 45 is i915. I started with removing i915 > > first. After a reboot the behavior of the system changed: the "no vector > > for"-message was gone. > > > > I added i915 again, rebooted, and the error was back. > > Could you please remove the i915 module and provide the output of > > lspci -vvv -s 00:02.0 Yet another thing which would be interesting. Load the i915 module and run the userland graphics stack. Open the serial port for read, which should not result in any interrupts. Check /proc/interrupts for the number of interrupts for the serial port, wait 10 seconds and check again. In theory the numbers should be the same. My crystal ball tells me they are not. Thanks, tglx
