-
Notifications
You must be signed in to change notification settings - Fork 4.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Random comparison difference in HLT/SiStrip/ControlView #41200
Comments
assign geometry |
New categories assigned: geometry @mdhildreth,@Dr15Jones,@makortel,@bsunanda,@civanch you have been requested to review this Pull request/Issue and eventually sign? Thanks |
A new Issue was created by @makortel Matti Kortelainen. @Dr15Jones, @perrotta, @dpiparo, @rappoccio, @makortel, @smuzaffar can you please review it and eventually sign/assign? Thanks. cms-bot commands are listed here |
is architecture dependence excluded (aka INTEL vs AMD)? |
Good point. I checked the PR test and baseline runTheMatrix output of #41186 (comment), and both were run on |
I'm running valgrind on step1 and see a bunch of these
need to be understood if related specifically to DD4HEP |
my valgrind command
|
not sure if the valgrind report is actually still related to this |
@VinInn , this issue was understood as a compiler bug when -O3 optimisation is used. The solution was to use -O2 optimisation for VecGeom. However, I am not sure if the problem Matti are reporting here is the same. |
I got the report above in latest 13_1_X nighty. Maybe understood, not solved apparently. |
one more valgrind message in step2 (most probably for a different issue)
|
Another occurrence in #41274 (comment), this time in workflow 12434.0 |
assign dqm |
New categories assigned: dqm @micsucmed,@rvenditti,@emanueleusai,@syuvivida,@pmandrik you have been requested to review this Pull request/Issue and eventually sign? Thanks |
FYI @cms-sw/trk-dpg-l2 |
Another occurrence in #41328 (comment), this time in workflow 12434.0 |
Another occurrence in #41460 (comment) in workflow 12434.0 |
Another occurrence in #41876 (comment) in workflow 12434.0 |
Another occurrence in cms-sw/cmsdist#8545 (comment) in workflow 12434.0 (although there because of an update of the compiler minor(?) differences in generated code can not be excluded) |
Another occurrence in #42075 (comment) in workflow 12434.0. |
type trk |
The
HLT/SiStrip/ControlView/{ClusterStoNCorr_OnTrack_FECCratevsFECSlot,ClusterStoNCorr_OnTrack_FECSlotVsFECRing_TECP}
histograms showed differences in workflow 11634.911 in PR tests of #41186 (comment) . The PR itself is very unlikely to be the cause of the differences. The differences have also not been visible in other recent PR tests, so these differences have likely random origin. The purpose of this issue is to nevertheless document them, in case they are visible in other tests later on.The 11634.911 is the DD4Hep workflow that, IIUC, reads the geometry from the XML file instead from the CondDB. These differences may be evidence of some rare non-reproducibility in DD4Hep code path (that we have observed, but not really solved, before).
The text was updated successfully, but these errors were encountered: