homepage

Ok I've verified that the patch does indeed fix the leak detected. Thank you very much INADA for knowing that there was a leak in the warnings module, I would have never guessed, especially given the tracemalloc stack given. Had it showed a callstack where the warning was created it would have made a lot more sense.
I agree this can be closed, however can the leak fix PLEASE be put into 3.6 (any any other version that needs it)? Who cares if warnings are 1.4x slower with the fix? Are you going to rationally tell me that keeping warnings fast is more important than fixing leaks? In most applications there should be no warnings so it doesn't really matter. This particular leak was causing our application to fail after running for a few days which makes it unusable in production. It's caused me a lot of days wasted in investigation. If speed was really a problem that would have been a much worthier thing to spend time on than finding leaks.
leaks should be highest priority, then speed. No rational developer would have complained that warnings got slower, that's when you fix warnings, not because of leaks! :)