Bug#951497: I can't reproduce

2020-02-24 Thread Matthew Fernandez
You’re definitely testing v2020.01.27-1 right, Adam? I would think you probably are, but I ask because the version you uploaded for me last week (v2020.02.17-1) is not the problem one. v2020.02.17-1 should not reproduce this issue as I also made the fix there before uploading it. The problems

Bug#951497: I can't reproduce

2020-02-24 Thread Adam Borowski
Hmm... it was me who reviewed+uploaded the current version, and it doesn't even put a strain on my box, despite using 4× as many cores but only 64GB memory. Ie, half the memory per thread. It also has an interesting pattern of CPU use: there's never any substantial load on any single thread, and