https://bugs.kde.org/show_bug.cgi?id=509428

--- Comment #2 from Arjen Hiemstra <[email protected]> ---
So the memory calculations were changed a while back, so now what we do is read
a "rough" value from `/proc/$PID/statm` and a "precise" value from
`/proc/$PID/smaps_rollup`, with the latter updating on a much more reduced
frequency (as its a CPU intensive operation to read that file). We then
calculate the difference between precise and rough value and when we read a new
value from `statm` we add the previously calculated difference to get at a more
"precise" value. I assumed these numbers would always result in a positive
number, but apparently you found a corner case where that's incorrect. Ensuring
it's at least 0 should be fairly trivial.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to