-
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
Exception when calling the method EventSetup module RPCConeBuilder #36786
Comments
A new Issue was created by @elfontan Elisa Fontanesi. @Dr15Jones, @perrotta, @dpiparo, @makortel, @smuzaffar, @qliphy can you please review it and eventually sign/assign? Thanks. cms-bot commands are listed here |
@bsunanda @cms-sw/geometry-l2 RPC geometry was changed in pre4, could this be the cause of this issue? |
Could you specify which workflow sees this exception? Please also put RPC experts in this link |
hi @bsunanda! the command to reproduce this is given by Elisa in the description:
|
FYI @cms-sw/rpc-dpg-l2 |
assign hlt I suspect that the issue may come from using as input file for |
New categories assigned: hlt @missirol,@Martin-Grunewald you have been requested to review this Pull request/Issue and eventually sign? Thanks |
Thanks for reporting, Elisa. I have tried to check a few things, and here's my current understanding.
So, I think HLT exposes the problem because of the use of Experts are free to correct any of the above, and can hopefully provide further feedback. |
assign geometry,l1 |
New categories assigned: geometry,l1 @cvuosalo,@mdhildreth,@epalencia,@ianna,@Dr15Jones,@rekovic,@makortel,@cecilecaillol,@civanch you have been requested to review this Pull request/Issue and eventually sign? Thanks |
Thank you @missirol for the detailed investigation |
urgent |
In CMSSW_12_3_0_pre4, I tried the command, but no error occurs.
How can the error be reproduced? |
You would need to run the extracted python file: However, the L1T xml file is not in the release itself:
@elfontan Please post complete instructions to reproduce! |
Hi @Martin-Grunewald, sure, here the instructions to use the new menu, sorry: cmsrel CMSSW_12_3_0_pre4
scram b -j 8 Thank you for the help, |
I assume in the above you mean |
Yes! |
Hi all, I am not an expert but in the code there is the check if the geometry has a complete ring. In the case of demo chambers, which are added the ring is not filled, only 2 sectors are added. May be these lines of the code should be commented out? |
Thank you @civanch , I think it makes sense. |
Hi @civanch, |
Great! Who will make PR with the fix? |
It will be nice if Elisa makes the PR
…________________________________
From: Vladimir Ivantchenko ***@***.***
Sent: 25 January 2022 15:04
To: cms-sw/cmssw
Cc: Sunanda Banerjee; Mention
Subject: Re: [cms-sw/cmssw] Exception when calling the method EventSetup module RPCConeBuilder (Issue #36786)
Great! Who will make PR with the fix?
—
Reply to this email directly, view it on GitHub<#36786 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABGMZOU5VAZXLUNTG7NHDMLUX2UWRANCNFSM5MVEEZWQ>.
Triage notifications on the go with GitHub Mobile for iOS<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
I can prepare the PR, sure. |
I had the same question above |
In this a bit stress situation I would not remove these lines but comment them out and add a comment in the code with reference to this issue. Code authors may improve the check later if needed. |
I tested by changing line 135 of of
to
This change prevents the exception, but then the program encounters a segmentation violation soon after this section of the code. I think 3014 and 4014 lack counter-rings. |
I'm puzzled. If we comment the check as a whole, the keys 3014 and 4014 pass as well, but they apparently do not originate a segmentation violation afterwards, see #36786 (comment) Can anybody explain? Carl, are you getting the same segmentation violation if you simply comment out the whole check? |
Hi all, The RPCConeBuilder.cc is a part of the L1Trigger/RPCTrigger, which is the emulator of the Run 1 RPC PAC Trigger, which is not used in the L1 Trigger decision since 2016. Adding @konec just in case. |
Maybe. As noted in #36786 (comment) [*], if I use the command in the original bug report, and work around the RPC problem with either an older GT or commenting out the exception, I do see a segmentation fault, but coming from a different L1T producer [**]. At first glance, this looks unrelated to the RPC issue at hand, but that's not to say it should not be debugged (I'm assuming this requires a look by L1T experts). I didn't catch in Carl's message if he thinks that 'his' segmentation violation comes from [*]
[**] Pseudo-recipe: # in 12_3_0_pre4, after commenting out the RPCConeBuilder exception
hltGetConfiguration /dev/CMSSW_12_3_0/GRun --mc --full --unprescale --globaltag 123X_mcRun3_2021_realistic_v4 \
--process HLTX --output minimal --l1-emulator uGT --max-events 1 \
--input root://xrootd-cms.infn.it//store/relval/CMSSW_12_0_1/RelValTTbar_14TeV/GEN-SIM-DIGI-RAW/120X_mcRun3_2021_realistic_v7-v2/10000/b67e121b-b29f-4eb0-8628-b3aa1cb76720.root \
> tmp.py && cmsRun tmp.py &> tmp.log Stack trace attached. Some maybe-relevant parts:
|
Hi @missirol , all, I would suggest then to introduce Carl's fix instead of simply comment the check and proceed (and I would include Karol's explanation as a description of the module). Let me know what you think. In case I can update the PR soon after. Cheers, [1] |
+1 |
+hlt #36803 fixed the problem at hand, but based on comments in this issue
|
Just for completeness, regarding the crash from
|
For completeness: a fix for the seg-fault discussed in #36786 (comment) and #36786 (comment) was provided by @elfontan in #36839. @perrotta Yes, this issue is resolved (Elisa should correct me if I'm wrong). Both problems came from L1T and Elisa provided fixes for those. I still find it sub-optimal that the L1T re-emulation for Run 3 brings in L1T ESProducers from Run 1 (see #36786 (comment)), but this goes beyond this issue. |
Thank you @missirol, I agree. |
|
Hi all,
while testing the new CMSSW_12_3_0_pre4 release to run the L1 emulation, I encountered the following error that I was not seeing before:
The mapping of bin lower bounds to indices does not contain all possible entries!!!
%MSG
----- Begin Fatal Exception 23-Jan-2022 16:52:55 CET-----------------------
An exception of category 'RPCInternal' occurred while
[0] Processing global begin Run run: 1
[1] Prefetching for module EventSetupRecordDataGetter/'hltGetConditions'
[2] Calling method for EventSetup module RPCConeBuilder/''
Exception Message:
Size differs for ring 3014 +- 100
----- End Fatal Exception -------------------------------------------------
Here the command that I used:
hltGetConfiguration /dev/CMSSW_12_3_0/GRun --mc --full --unprescale --globaltag auto:phase1_2021_realistic --process MYHLT --output minimal --l1Xml L1Menu_Collisions2022_v0_1_2.xml --l1-emulator uGT --input root://xrootd-cms.infn.it//store/relval/CMSSW_12_0_1/RelValTTbar_14TeV/GEN-SIM-DIGI-RAW/120X_mcRun3_2021_realistic_v7-v2/10000/b67e121b-b29f-4eb0-8628-b3aa1cb76720.root >& hlt_uGT.py
Does anyone have an idea of the reason?
Thank you very much in advance,
--Elisa
The text was updated successfully, but these errors were encountered: