Dan Riley

> ```diff > + printf("wordGPU %d %d %d \n",i, ww, rawId); > ``` Shouldn't that be ``` printf("wordGPU %ud %ud %ud \n",i, ww, rawId); ``` I'm surprised the compiler didn't...

> @dan131riley as far as I can read from the description, this PR is still a work in progress, as there are things to be understood and bugs to be...

@fwyzard and @slava77 thanks for the careful reviews. wrt > should there be a matrix workflow (some `.50X`) to regularly test this code? yes there should be, will address the...

@slava77 It is being worked on, should have an update early next week.

Some context, seems to be a lot going wrong (charge non-conservation!!): ``` %MSG-w LogicError: LheWeightValidation:lheWeightValidation@beginRun 08-Mar-2022 06:38:42 CET Run: 1 ::getByLabel: An attempt was made to read a Run product...

Collecting context: ``` %MSG-e HepMCValidationHelper: MBUEandQCDValidation:mbueAndqcdValidation 11-Apr-2022 05:03:21 CEST Run: 1 Event: 1 Tau with no daughters. This is a bug. Fix it %MSG A fatal system signal has occurred:...

Not seeing much of a pattern here: ``` %MSG-e HepMCValidationHelper: MBUEandQCDValidation:mbueAndqcdValidation 13-Apr-2022 05:51:50 CEST Run: 1 Event: 1 Tau with no daughters. This is a bug. Fix it %MSG A...

> > The floating point exception ROOT error messages are suspicious. Are we having two signal handlers being run here? > > That is really weird. We disable ROOT's signal...

@Dr15Jones found some discussion from three years ago at https://github.com/cms-sw/cmssw/pull/28112#issuecomment-539173390 This looks like the same thing, we don't catch SIGFPE and something is enabling FPE's behind our backs. Then the...

Thinking about this some more, if something were turning on FPEs globally then I'd expect to get a ton of crashes, since we generate lots of NaNs and junk in...