W. David Dagenhart

Just randomly looking for a correct spin lock and I stumbled on this one first, I was going to quote this one as the correct way to do it, but...

This one looks correct. https://cmssdt.cern.ch/lxr/source/FWCore/Services/plugins/ConcurrentModuleTimer.cc#0206

Just skimming through looking for obvious spin locks with compare_exchange_strong, all the rest of them look OK. Just luck that the first other one I looked at was bad... There...

I could easily fix the two spin locks to work correctly. Ask and I'll do that. I suspect that will fix the problem... That is what I would suggest. We...

atomic_flag should also work. https://en.cppreference.com/w/cpp/atomic/atomic_flag If you don't use the C++20 extension, the interface looks simpler for a simple spin lock, maybe less prone to error. Guaranteed to be lock...

One thing is worth noting before we continue. In the use case where one lumisection ends (lets call it N) and another starts (call it N+1) and there is only...

I think the focus of discussion in Mattermost was on the lumiQueue_ that enforces the maximum number of luminosity blocks that can be processed concurrently. I don't think that is...

Most of my comments above have been discussing item 2 listed at the top of this Issue. I do not understand item 1. I am not aware of any hard...

I've read through the last post several times. I do not understand what is going on. The only theory we have so far as to why lumi N is delayed...

There is a new unit test that reproduces the first item described in the comment at the top of this Issue. See pull request #43018. It is not a fix...