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
SiStrip (un)packer: fixes and support for nonstandard ZS(lite) modes #23417
SiStrip (un)packer: fixes and support for nonstandard ZS(lite) modes #23417
Conversation
- rename the PACKET_CODE definitions for ZS lite to ZS - change the packetCode implementation: default (0) for ZS lite modes, taken from the third byte for ZS Reason: for ZS lite there are no packet codes, the packing scheme is given by the readout mode; for non-lite ZS it is in the channel data
- add "PacketCode" option to SiStripDigiToRawModule - add packetCodeFromString conversion in SiStripFEDBufferComponents - pass the packet code on to FEDBufferPayloadCreator::fillChannelBuffer
…saggio/cmssw into SiStripUnpacker_fixZSmodes_102
The code-checks are being triggered in jenkins. |
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-23417/4957 |
A new Pull Request was created by @pieterdavid (Pieter David) for master. It involves the following packages: EventFilter/SiStripRawToDigi @perrotta, @cmsbuild, @slava77 can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
@icali FYI |
@cmsbuild please test |
Thanks @perrotta , that message (and the corresponding summary of unpacker warnings at the end of the job) is indeed present in the three jobs with differences. I will investigate |
+1
|
+1 |
Output of the packer-unpacker closure test (run with this script), as requested during Friday's RECO meeting:
|
On the "FED data missing" warnings (likely due to disabled FEDs): for the 1000.0 and 1001.0 workflows (run 165121) they are all about FED 87. For the 4.22 workflow (2011 cosmics, run 160960) the warning is there for three FEDs, including 87 (I tried to rerun with debugging on to get the other two IDs, but the das query doesn't return any results, and if I take the file list from the comparisons the cern xrootd server says that the file doesn't exist - indeed I do not see it on eos). |
The other two missing FED IDs should be 103 and 388 which are two FEDs that have been excluded for a long time already, FED 87 was a transitory problem that so happened to be present in the run that is used. |
for testing those workflows you can use the
see discussion at #22278 |
Thank you all for checking |
Hi @pieterdavid , thanks for the further checks. and in my recollection they are not permanently out of the cabling. Indeed looks like a transient problem (also in this case both the FEDs are seen in the offline masking for the |
@pieterdavid @perrotta from the thread above I understand that the new messages appearing for 2011 data are related to a particular data period and understood, and should not represent a general problem of the PR. In any case, I consider that the legacy release to study Run1 remains the 53X. |
+operations the updates to the StandardSequences appear coherent with the rest of the PR |
This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @davidlange6, @slava77, @smuzaffar, @fabiocos (and backports should be raised in the release meeting by the corresponding L2) |
Yes @fabiocos , I confirm that the messages are only exposed by this PR (as new warning messages are issued which did not exist before unless in some debug mode). |
+1 |
This PR contains fixes for some bugs in the SiStrip unpacker for 10bit ZS(lite) readout modes (found by @erikbutz when testing the new hybrid zero suppression for heavy ion data taking later this year), and adds support for all different packing schemes (readout modes and packet codes, many were missing) to the packer (used in simulation and to repack HI data in the HLT).
The unpacker changes are limited to the code used for 10bit packed data (plus the code in the EDProducer to add cases for the modes that were not taken into account there), so there should be no effect on what is used for normal pp data. The packer changes were tested by unpacking its output and comparing the digis to the original collections, taking into account the expected differences (see this config and related code).
with @alesaggio
CC: @OlivierBondu @vidalm @mmusich @echabert