memory/info has started spewing invalid data on passing results

[Jeff Lane ]

Ive been seeing a LOT of memory/info results that look like the following:

PASS: Difference is -4014804992 bytes (-93.48%) and less than the 10% threshold.
 /proc/meminfo reports: 8309772288 kB
 dmi reports: 4294967296 kB

At some point in recent history, we started getting negative numbers in the meminfo calculations and the script itself has no error handling to catch this sort of mistake.

To fix:
A: Figure out if the negative numbers are because of an output change in the files that the script looks at or an error in the script itself.
B: Add error checking so the script complains when it DOES encounter negative amounts.

[Jeff Lane  (bladernr)]

https://certification.canonical.com/hardware/201206-11410/submission/d57tJbS6jlkibtY/result/test:memory%252Finfo:__memory__

[Brendan Donegan (brendan-donegan)]

Already fixed under a different bug number, so marking this as a duplicate even though it's older