March 3, 20179 yr Hi Not really sure what's going on here to be honest, but it says I should post and ask for help - help please! Should I be worried? Thanks in advance highlander-diagnostics-20170303-0006.zip
March 3, 20179 yr Somebody a little more familiar with the modules is going to have to pipe in. Here's the interrupt in question: Quote Mar 2 23:40:26 Highlander root: 16: ..... IR-IO-APIC 16-fasteoi vfio-intx(0000:0b:00.0) To let you know whether this is harmless or not. (I just write the tests -> Not an expert on everything )
March 4, 20179 yr On 3/2/2017 at 7:55 PM, Squid said: Somebody a little more familiar with the modules is going to have to pipe in. Here's the interrupt in question: Quote Mar 2 23:40:26 Highlander root: 16: ..... IR-IO-APIC 16-fasteoi vfio-intx(0000:0b:00.0) To let you know whether this is harmless or not. (I just write the tests -> Not an expert on everything ) I looked at it earlier, and again just now, and it's confusing. I can't say I understand what happened. For one thing, the message "irq 16: nobody cared" occurred twice, and IRQ 16 was disabled twice! Never seen that happen before. It means that either IRQ 16 was re-enabled (don't think that ever happens) or the kernel forgot it had disabled it previously, and had allowed processes to re-hook it. The PCI address corresponds to your Hauppauge TV Tuner, the first one on the card. I assume that means this is a DVB build. Since an unhandled IRQ is a bug, in some piece of related software, all you can do is make sure you have the latest motherboard BIOS, latest firmware for the card, latest driver for the card, and the latest kernel. Something else that is confusing (and I don't see how it can be related), *both* occurrences of this issue happened apparently just as the appdata backup started, one second before the first Docker stopped. I don't know how it could be related, but it seems too close in time to rule out a connection. Also probably unrelated, but what is also confusing to me is that the Docker update seems to occur just before this, and then also during this appdata backup. Perhaps Squid can explain... (or I'm reading the syslog line items wrong).
March 4, 20179 yr Author 7 hours ago, RobJ said: I looked at it earlier, and again just now, and it's confusing. I can't say I understand what happened. For one thing, the message "irq 16: nobody cared" occurred twice, and IRQ 16 was disabled twice! Never seen that happen before. It means that either IRQ 16 was re-enabled (don't think that ever happens) or the kernel forgot it had disabled it previously, and had allowed processes to re-hook it. The PCI address corresponds to your Hauppauge TV Tuner, the first one on the card. I assume that means this is a DVB build. Since an unhandled IRQ is a bug, in some piece of related software, all you can do is make sure you have the latest motherboard BIOS, latest firmware for the card, latest driver for the card, and the latest kernel. Something else that is confusing (and I don't see how it can be related), *both* occurrences of this issue happened apparently just as the appdata backup started, one second before the first Docker stopped. I don't know how it could be related, but it seems too close in time to rule out a connection. Also probably unrelated, but what is also confusing to me is that the Docker update seems to occur just before this, and then also during this appdata backup. Perhaps Squid can explain... (or I'm reading the syslog line items wrong). Thanks RobJ. My TV card was playing up (stuttering) so that could be related. I think you've given me a clue with the appdata insight. I have a tvheadend docker I don't use as I use dvblink in a VM with the TV card passed through. The tvheadend docker isn't running, but it still has the extra paremeters line to passthrough the controllers - maybe this is confusing things? To be safe, I'll remove tvheadend from the appdata backup as it's not vital.
Archived
This topic is now archived and is closed to further replies.