Matej Roguljic
Matej Roguljic
Per cmsbot, this accounts for roughly half the differences observed in the main PR. I've ran valgrind on one event in step 3 of 140.045 in two environments: reference (with...
^ testing to check the number of diffs when we have the `assert`, but with an explicit initializazion and (dummy) call to `ygx0_` members `ygsig`. They are unused and uninitialized...
^ testing to see if declaration of variables at initialization will stabilize the output (as suggested [here](https://github.com/cms-sw/cmssw/pull/48356#issuecomment-2996517864))
Apologies, I missed the previous comment. The only difference is the removal of "RecoGlobalFakeHLT" step in the related worfkflow. I believe it is not defined in 150X, and runTheMatrix would...
> please test for CMSSW_14_2_UBSAN_X I just ran some local tests for 29634.402 and it seems to hang, will have to do further checks