Ticket #4527 (closed: fixed)
Spurious crash during ConvertToDiffractionMDWorkspace
| Reported by: | Owen Arnold | Owned by: | Janik Zikovsky |
|---|---|---|---|
| Priority: | critical | Milestone: | Release 2.0 |
| Component: | Mantid | Keywords: | |
| Cc: | Pascal.Manuel@… | Blocked By: | |
| Blocking: | Tester: | Owen Arnold |
Description
Has been identified on Ubuntu 10.4 by Pascal, but may affect other environments.
Load the SXD data (SXD23767.RAW). Run ConvertToDiffractionMD with Q lab. This will sometimes crash MantidPlot, and sometimes run without issue. Do we have a race condition issue here?
Change History
comment:3 Changed 9 years ago by Owen Arnold
From Pascal:
Actually, I can make it crash even without doing an instrument view before
from mantidsimple import * LoadRaw(Filename='/home/pm/Science/WISH_commissioning/Data/archive/cycle_11_1/WISH00017946.raw',OutputWorkspace='17946',LoadLogFiles='0',LoadMonitors='Exclude') LoadIsawUB(InputWorkspace='17946',Filename=r'/home/pm/Desktop/SrEr2O4_ISAW_UB.mat') ConvertToDiffractionMDWorkspace(InputWorkspace='17946',OutputWorkspace='17946_MD_HKL',OutputDimensions='HKL') BinToMDHistoWorkspace(InputWorkspace='17946_MD_HKL',AlignedDimX='H,-1.0,8.0,200',AlignedDimY='K,-1.0,8.0,200',AlignedDimZ='L,0,1.5,200',OutputWorkspace='test_rebin') Then click on L->Y, K->X and try to do a cut.
comment:4 Changed 9 years ago by Owen Arnold
Pascal has provided a demo of the crash:
http://www.fileconvoy.com/dfl.php?id=g16aae3619357eba6862258d0aaf5ec1782249
comment:5 Changed 9 years ago by Owen Arnold
Sorry, Last two items should actually form a separate ticket #4543.
comment:6 Changed 9 years ago by Janik Zikovsky
Looks like it has to do with the dummy detectors that SXD has. They are not being caught correctly...
comment:7 Changed 9 years ago by Janik Zikovsky
Created a new ticket for the crash: #4544 which is in logging. But I can work around it easily.
comment:8 Changed 9 years ago by Janik Zikovsky
Refs #4527 fix crash due to non-thread safe logging
Changeset: d8213ff150b962e0c745b542182c3f3a89aa993a
comment:9 Changed 9 years ago by Janik Zikovsky
- Status changed from accepted to verify
- Resolution set to fixed
comment:10 Changed 9 years ago by Owen Arnold
- Status changed from verify to verifying
- Tester set to Owen Arnold
comment:11 Changed 9 years ago by Owen Arnold
Since I wasn't able to replicate this issue on my environment. I'm going to wait to hear back from Pascal about it. I've suggested that he generates a loop in python and runs the algorithms n-times to try to trigger the thread safety problem.
comment:12 Changed 9 years ago by Janik Zikovsky
Refs #4527 fix crash due to non-thread safe logging
Changeset: d8213ff150b962e0c745b542182c3f3a89aa993a
comment:13 Changed 9 years ago by Owen Arnold
- Status changed from verifying to closed
Pascal has tried this out several times since the fix and hasn't observed any crashes. I've tried it out as part of a python loop 10 times and couldn't find any unexpected behaviour.
comment:14 Changed 5 years ago by Stuart Campbell
This ticket has been transferred to github issue 5374
Janik, I've emailed you the data you need to reproduce this.