SWUTILS-959: Add AER error information

[dchadwic-xilinx]

This is a draft. Ultimately, there's some functionality I've added in the unmerged #31 and even #25 I'd like to pull in before this is done.

The output looks like this:

Information about the files we are monitoring: https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-bus-pci-devices-aer_stats

EDIT: Output now looks like this:

[dchadwic-xilinx]

I think it's probably worth getting the 'fatal' AER errors highlighted an an "interesting value".

[dchadwic-xilinx]

This table is now hidden by default and also is passed through the interesting values check. I temporarily changed the check for non-fatal to look for ==0 instead for the pics.

[amitgu-xilinx]

@dchadwic-xilinx Thank you, looks good. Joby reported that the AER error he observed, but the counters were not incremented. You might need to check with him.

Also you mentioned earlier that "rasdaemon" service must be running to collect the logs using PCI.
If so, have you tried the following way to read the AER counters?
"cat /sys/bus/pci/devices/0000:42:00.0/aer_dev_*" maha shared the following command to check AER stats.

[ibeecraft-amd]

I would like the interesting values sections to be a little better at highlighting either interface names or PCI IDs but as it has always been this way, I'm happy for this change to go in.

Maybe later we can modify the interesting values section.

[amitgu-xilinx]

As we discussed, please also report the PCIe device address in the " interesting values" section.

[dchadwic-xilinx]

I would like the interesting values sections to be a little better at highlighting either interface names or PCI IDs but as it has always been this way, I'm happy for this change to go in.

I've scoped this and it won't be so simple with the current way that tabulate() works. By the time it's comparing values, it doesn't have the context anymore. This is something that could change but this PR should not be blocked by it.

[dchadwic-xilinx]

By the time it's comparing values, it doesn't have the context anymore.

That being said - I've found a solution which pipes it though and gives this output:

@amitgu-xilinx @ibeecraft-amd does this meet your requirement? If so - I'll update the PR.

[dchadwic-xilinx]

This has also been improved to grab the full list of error numbers: