Massive memory usage by CTracer

[nedbat]

Originally reported by Brett Cannon (Bitbucket: brettcannon, GitHub: brettcannon)

---

When I run coverage over Python's stdlib (this is directly, not using allcoverage since I forgot to use the script), I end up w/ nearly 34 GB of disk space used. Now I really doubt that the .coverage file is using that much space as it was not that big at the end of the run.

So my hunch is that there is a fundamental memory leak that is leading to a massive amount of swap being put on disk. This happens on both OS X Lion and OS X Snow Leopard. When I run the coverage again using the pure Python code then it never seems to trigger a use of disk (other than writing to .coverage).

---

• Bitbucket: https://bitbucket.org/ned/coveragepy/issue/147

[nedbat]

Can you see where the disk space is going? There have been memory leaks in CTracer in the past, but not for a while, and others aren't reporting a problem. Perhaps fullcoverage torques it in some unusual way?

[nedbat]

Original comment by Brett Cannon (Bitbucket: brettcannon, GitHub: brettcannon)

---

I'll see what I can do, but FYI fullcoverage was not used to trigger this; ./python.exe -m coverage --pylib Lib/test/regrtest.py.

And just to get as much info recorded as possible, the memory hit starts with test_argparse which (I think) is the first pure Python module to be tested. And this is against CPython default so it could be a bug in CPython itself. I also wonder if this might be Python 3 related as I assume hardly anyone uses coverage.py against Python 3 quite yet.

[nedbat]

Fixed in <<changeset 7b050ee1103d (bb)>>.