coverage-linux workflow going out of memory

[targos]

https://github.com/nodejs/node/runs/3103690697?check_suite_focus=true
https://github.com/nodejs/node/runs/3102905499?check_suite_focus=true

[targos]

https://github.com/nodejs/node/runs/3231046711?check_suite_focus=true

[targos]

https://github.com/nodejs/node/runs/3238780429?check_suite_focus=true

@bcoe

[bcoe]

I think perhaps we just need to pass a larger --max-old-space-size=8192 to c8, it merges 1000s of coverage reports together, from each subprocess, and my hunch is that we're bumping into memory limits for the combined object.

[bcoe]

@targos I believe we've fixed this?