Possibly improve parsing OSRELEASE, etc

[lathiat]

I noticed today that you can use eu-readelf to parse the vmcoreinfo (OSRELEASE, BUILD-ID, PAGESIZE, etc) from this article:
https://blogs.oracle.com/linux/post/whats-inside-a-linux-kernel-core-dump

So we may be able to use either that or a python wrapper for elfutils etc to parse this data better than we currently do

[xmkg]

So I tried this eu-readelf thing, and it looks like it works on pure core files instead of kdump/compressed kdump core files, which means we'll need to locate the offset before calling the tool. I might be missing something, so please feel free to fill me in on how we can utilize this command on a raw kdump dump file.

The current implementation works exactly like the file command so I think it's reliable and fast enough for our purposes, but I would be eager to consider an alternative if it's more straightforward than what we currently have.

[xmkg]

Now, with #68 and #72 in place, the parser no longer relies on guesswork, and it properly parses the header, and the VMCOREINFO when it is present.

@lathiat do these two address the issue?