Unless you can provide more evidence, I'm marking this bug as invalid.
For other reasons I'm been watching /tmp content while selftest is running and obviously we clean it up as we go.
I have occurrences where /tmp is filling up GBs for *one* test, but it's related to thread and socket leaks (tracked elsewhere).
Unless you can provide more evidence, I'm marking this bug as invalid.
For other reasons I'm been watching /tmp content while selftest is running and obviously we clean it up as we go.
I have occurrences where /tmp is filling up GBs for *one* test, but it's related to thread and socket leaks (tracked elsewhere).